Physiological plasticity and local adaptation to elevated pCO 2 in calcareous algae: an ontogenetic and geographic approach

Abstract To project how ocean acidification will impact biological communities in the future, it is critical to understand the potential for local adaptation and the physiological plasticity of marine organisms throughout their entire life cycle, as some stages may be more vulnerable than others. Coralline algae are ecosystem engineers that play significant functional roles in oceans worldwide and are considered vulnerable to ocean acidification. Using different stages of coralline algae, we tested the hypothesis that populations living in environments with higher environmental variability and exposed to higher levels of pCO 2 would be less affected by high pCO 2 than populations from a more stable environment experiencing lower levels of pCO 2. Our results show that spores are less sensitive to elevated pCO 2 than adults. Spore growth and mortality were not affected by pCO 2 level; however, elevated pCO 2 negatively impacted the physiology and growth rates of adults, with stronger effects in populations that experienced both lower levels of pCO 2 and lower variability in carbonate chemistry, suggesting local adaptation. Differences in physiological plasticity and the potential for adaptation could have important implications for the ecological and evolutionary responses of coralline algae to future environmental changes.

1974; Morse, Andersson, & Mackenzie, 2006). This study compared the responses of different life-history stages of two populations of the articulate coralline algae Corallina vancouveriensis to different levels of pCO 2 .
This alga is an abundant species along the intertidal coast of the California Current Large Marine Ecosystem (CCLME) which is one of the most productive and economically important ecosystems on Earth (Costanza et al., 1997). This ecosystem experiences high variability in water chemistry due to upwelling events and is particularly sensitive to ocean acidification and global warming (Gruber et al., 2012;Hauri et al., 2009).
Algal distribution is, in part, the result of adaptive responses to longand short-term fluctuations in the environment. Thus, understanding the degree of phenotypic flexibility and local adaptation is essential for predicting changes in their biogeographic distributions and to project future ecological trends in response to global changes. Furthermore, a complete ecophysiological understanding that includes multiple life-history stages will help us to link physiological responses with fluctuations in the environment and to identify thresholds and vulnerabilities across the life cycle (Harley et al., 2012). Ocean acidification can impact physiological processes in algae such as photosynthesis, respiration, and growth, which are metabolically linked and can influence each other (Borowitzka et al., 1974;Gao et al., 1993;Martin, Charnoz, & Gattuso, 2013;Martin, Cohu, Vignot, Zimmerman, & Gattuso, 2013). CO 2 enrichment can stimulate growth and photosynthesis by providing more substrate for carbon fixation; however, some species of algae have carbon concentration mechanisms (CCMs) that facilitate the acquisition of carbon from other sources (Giordano, Beardall, & Raven, 2005;Raven, Giordano, Beardall, & Maberly, 2012). In the genus Corallina, algae have evolved CCMs that allow them to transform HCO − 3 (which is very abundant in the ocean) into CO 2 and thus are not carbon-limited. Species that do not possess CCMs are generally carbon-limited under current concentration of seawater CO 2 and thus are more likely to respond positively to elevated pCO 2 (Kubler, Johnston, & Raven, 1999). Thus, algal responses to pCO 2 will depend, in part, on the availability of carbon sources and the mechanisms present to obtain them.
The physiological response of calcifying algae to ocean acidification is highly variable, most likely reflecting the high diversity in this group, variation in photosynthetic pathways and calcification mechanisms, and variation in acclimatization capacity of different species (Koch et al., 2013).
Currently, relatively little is known about the effects of ocean acidification on early life-history stages of coralline algae (Bradassi, Cumani, Bressan, & Dupont, 2013;Cumani, Bradassi, Di Pascoli, & Bressan, 2010;Kroeker et al., 2012;Kuffner et al., 2007;Roleda et al., 2015) or about the capacity for local adaptation in this important group. Only two studies have looked at spore development of crustose coralline algae under ocean acidification (Bradassi et al., 2013;Cumani et al., 2010), and to our knowledge, only one study has looked at the effect of ocean acidification on spore development in articulate coralline algae (Roleda et al., 2015).
Corallina vancouveriensis reproduce by releasing spores (Johansen, 1981), which can fully attach to the bottom within hours of release (Miklasz, 2012) and recruit near the parental alga. The capacity of coralline algae to attach rapidly could limit dispersal distance, restrict gene flow among populations, and increase the potential for local adaptation in this species (Endler, 1977). Local adaptation can produce differences in physiology and life history and provide advantages in fitness in the local environment. Distinguishing spatial patterns of local adaptation and the relative contribution of local adaptation and phenotypic plasticity to organismal performance will help us to understand and predict the impacts of climate change and implement effective practices to manage marine ecosystems.
To explore the role of local adaptation and whether differences in physiological responses to high pCO 2 are consistent with regional differences in carbonate chemistry patterns, we cultured spores and adults of C. vancouveriensis from different populations. By measuring survival, growth, photosynthesis, and other physiological parameters, we explored the tolerance of different life stages to high pCO 2 , and whether different populations were locally adapted to environments with different pCO 2 levels. We hypothesized that populations of C. vancouveriensis living in environments with higher environmental variability (due to upwelling) and exposed to higher levels of pCO 2 would be less affected by similar high pCO 2 levels than populations from more stable environments experiencing lower pCO 2 levels.

| Algal collections and sensor deployment
Corallina vancouveriensis Yendo (1902) is a common articulate coralline algae in the CCLME. It is light pink to light purple in color and can form dense mats on emergent bedrock or in tidepools in mid-to-low intertidal zones of exposed habitats.  (Cudaback, 2005). Sites north of Point Conception experience high wave exposure and waters are around 3-4°C colder due to higher recurrence of coastal upwelling (Blanchette, Miner, & Gaines, 2002), whereas south of Point Conception waters are warmer and the shore is more protected from heavy wave action (O'Reilly & Guza, 1993).
Algae were individually stored in plastic bags, placed in a cooler, and transported to the University of California Santa Barbara. In the laboratory, algal fronds were thoroughly and gently cleaned of epiphytic organisms and accumulated sediments and placed in tanks with running seawater at 14-15°C. Healthy algal specimens, that is, without alteration of the cortical tissue or discoloration, were selected to Temperature and pH sensors were deployed north and south of Point Conception in 2013, as a part of a larger network of sensors making continuous measurements at sites throughout the CCLME (Hofmann et al., 2014, Chan et al. in prep). Temperature and pH were measured every 20 min using Durafet ® -based (Honeywell Inc.) pH sensors that were custom-designed for near-shore deployment (Chan et al. in prep). Sensors were secured to the bedrock and placed submerged in tide pools in the lower part of the intertidal zone. Unfortunately, sensors at our collection sites failed to record pCO 2 . Thus, pCO 2 data were obtained from the closest network sensors to our collection sites (34°28.03N, 120°16.69W north of Point Conception and 34°43.14N, 120°36.53 south of Point Conception). Spearman correlation analyses were conducted to compare the environments (temperature) at the sites where the sensors were located and the sites where the algal collections were performed.

| Algal culturing and seawater chemistry
Adults and spores of C. vancouveriensis were cultured under different pCO 2 levels using a flow-through CO 2 mixing system as described in Fangue et al. (2010). The system blends dry, CO 2 -free atmospheric air with pure CO 2 to produce different pCO 2 levels using mass flow controllers. Gas for each mixture was continually delivered to gas-mixing reservoirs for equilibration with seawater to achieve a desired pCO 2 level.
One header tank was used for each pCO 2 treatment, and each treatment had two experimental tanks that were randomized with interdependent replicates within treatments . CO 2 -equilibrated seawater was then transferred from the reservoir buckets to the larval buckets for the duration of the experiment using lawn irrigation drippers.
The system was modified by replacing culture buckets with rectangular tanks and adding LED lights overhead (MarineLand Reef). Small submersible aquarium pumps (Aquatop, 70 gph) and pipes were used to provide uniform water flow inside the tanks, with a flow rate ~1 cm/s. Two pCO 2 levels were compared: for adults ~410 μatm (pH = 8.0) and 1,033 μatm (pH = 7.7); for spores ~485 (pH = 8.0) and 1,186 μatm (pH = 7.6) ( Temperature, salinity, and pH were measured daily for each pCO 2 experimental treatment according to best-practice procedures (Dickson, Sabine, & Christian, 2007;Fangue et al., 2010). Temperature was measured using a wire thermocouple (Themolyne PM 207000/Series 1218), and salinity was measured using a conductivity meter (YSI 3100). pH was determined following the standard operating procedure (SOP) 6b (Dickson et al., 2007) using a spectrophotometer (Bio Spec-1601; Shimadzu) and dye m-cresol purple (Sigma-Aldrich) as the indicator. Total alkalinity (TA) was measured every 3 days in the reservoir buckets, following the SOP 3b (Dickson et al., 2007). Water samples for TA were collected using borosilicate glass-stoppered bottles, poisoned with mercuric chloride, and analyzed at a later time using a potentiometric titration procedure with a commercially available titration unit (T50; Mettler Toledo) and following the SOP 3b (Dickson et al., 2007). Both pH and alkalinity were assessed for accuracy using certified reference materials ( (Gao et al., 1993). Temperature, salinity, and carbonate parameters of seawater used in experimental treatments are shown in Table 1.

| Spore and crust (juvenile) growth and mortality
Within 3 days after collection, a subset of algal specimens was haphazardly selected to obtain spores. Five or six fronds (~6-7 cm) per individual were placed on previously labeled cover glass slides in a 300-ml container filled with filtered seawater. Lids were placed on containers, and fronds were left to release spores naturally for 1 day at room temperature (19-20°C). After 24 hr, the cover glass slides with spores were transferred to the experimental tanks (high and low pCO 2 treatments in duplicate) and cultured for 19 days. Each experimental unit received slides with spores from the four different sites (n = 15-17 for Santa Barbara, n = 14-16 for Carpinteria, n = 5-7 for Cambria, and n = 8-10 for Arroyo Grande). The number of slides per site was dependent on the amount of spore material released by the algae: n = 19, 14, 8, 10 (high pCO 2 tank 1); n = 18, 14, 5, 8 (high pCO 2 tank 2); n = 23, 16, 7, 7 (low pCO 2 tank 1); n = 15, 14, 7, 11 (low pCO 2 tank 2), Santa Barbara, Carpinteria, Cambria, and Arroyo Grande, respectively.
Spore growth and mortality were monitored and recorded under both low and high pCO 2 conditions at 3 and 19 days after settlement using a dissecting scope and a digital camera (Jenoptik). Growth rates were estimated by measuring crust surface area over time using ImageJ software ver. 1.42 (Abramoff, Magalhaes, & Ram, 2004). Photographs were taken under 8× magnification, using a grid under the glass slide to ensure that the same spores were photographed every time.

| Adult physiology and growth
From each site, 12 adult algae (n = 48) were selected for physiological analyses. Six young branches (~1.5 cm and 100-120 mg fresh weight) were excised from each individual and randomly assigned to the experimental treatments. These branches were inserted upright into plastic grids at the bottom of experimental tanks and cultured for 30 days at 14 ± 1°C under both low (~410 μatm) and high pCO 2 (~1030 μatm) conditions (Table 1).

| Metabolic rates and primary productivity
Net primary productivity (NPP) and respiration (R) were measured at the beginning (Day 0) and at the end (Day 30) of the experimental treatments using the light and dark bottle methodology described in Howarth and Michaels (2000). These physiological rates were assessed as changes in dissolved oxygen concentrations during light and dark incubations, respectively. In brief, branches were placed in 50-ml acrylic chambers filled with seawater at the same pCO 2 as the experimental tank that was previously filtered and sterilized with UV light. Metabolic chambers were kept in a temperature-and light-controlled incubation tanks (30 ± 2.5 μmol photon m −2 s −1 and 14 ± 1°C) for 3 hr in order to avoid oxygen saturation greater than 120% during light incubation and to maintain oxygen saturation above 80% at the end of the dark incubation (Noisette, Egilsdottir, Davoult, & Martin, 2013

| Growth rates and biochemical components
Vegetative growth of each algae branch was determined by changes in the wet weight between the beginning and the end of the experiment.
The relative growth rate (RGR), expressed as percentage increase in fresh weight biomass per day (%/day), was estimated assuming exponential growth during the culture period according to the formula: where W 0 represents the initial and W t the final wet weight of the algae, and t is the time of culture in days.
Photosynthetic pigments were measured following Gao and Zheng (2010). About 0.1 g FW per sample was ground and placed in 10 ml absolute methanol at 4°C in darkness for 24 hr. Chlorophyll a and carotenoids were determined spectrophotometrically according to Wellburn (1994). For phycobiliproteins (i.e. phycocyanin and phycoerythrin), samples of about 0.1 g FW were placed in 5 ml of 0.1 M phosphate buffer (pH 6.8), ground at 4°C, and rinsed with a further 5 ml of buffer for 24 hr. The concentrations of phycobiliproteins were measured spectrophotometrically using the chromatic equations of (Beer & Eshel, 1985). All pigments were measured using the spectrophotometer Bio Spec-1601 (Shimadzu) after centrifugation at 5,000 g for 15 min.

| RESULTS
We collected algae from four sites, which consistently experience different strengths and durations of upwelling events (Blanchette et al., 2002). Sensors deployed in winter 2013 revealed strong differences in temperature and carbonate chemistry between intertidal sites located north and south of Point Conception (Fig. 1). At the northern site, pCO 2 reached a maximum of 2,904 μatm (pH = 7.24) and varied around a mean pH of 7.45 ± 0.13 and a mean pCO 2 of 1,770 ± 450 μatm, whereas at the southern site, pCO 2 reached a maximum of 946.6 μatm (pH = 7.7) and varied around a mean pH of 8.14 ± 0.2 and a mean pCO 2 of 350 ± 189 μatm. Correlations of temperature time series showed a strong positive relationship between collections sites and the sensor sites both north and south of Point Conception (Spearman correlation coefficient for northern sites ρ = .705, p < .001 and southern sites ρ = .631, p < .001, Fig. S1).

| Spore growth and mortality
Crust growth differed significantly among the four sites (Fig. 2a, areas at the end of the experiment (only 52.25% increase). We found that spore growth in response to high pCO 2 differed among populations, with higher growth rates in Santa Barbara compared to the other populations (Tukey's multiple comparison, p < .05; Fig. 2a, Table 2).
Although crusts from Arroyo Grande, Cambria, and Carpinteria showed slightly lower growth under high pCO 2 (Fig. 2a), our Tukey's comparison test revealed that these differences were not statistically significant. Mortality of spores differed among sites (Table 2, Fig. 2) but did not differ between pCO 2 treatments (Fig. 2b, Table 2). These differences were explained by the lowest proportion of dead spores in Carpinteria (southern population) compared to the other sites (Tukey's multiple comparison, p < .05).

| Adult physiology and growth
Net primary productivity, GPP, and respiration rates of C. vancouveriensis differed greatly depending on pCO 2 level and the population of origin (Figs 3, S2,  Fig. S1). In contrast to photosynthetic rates, respiratory rates increased for all sites after culturing the algae at high pCO 2 levels for 30 days (F 3,88 = 3.5, p = .018, and Tukey's multiple comparison, p < .001, Fig. 3), while respiration remained unchanged in the low pCO 2 treatment after 30 days (p > .05, Fig. 3).
Relative growth rates of C. vancouveriensis showed a significant interaction between pCO 2 treatment and site (F 3,184 = 8.78, p < .001, Fig. 4). Overall, RGR did not differ between populations under low pCO 2 (Tukey's multiple comparison, p > .05, Fig. 4)    pCO 2 levels (southern sites) were more sensitive to the highest tested pCO 2 . Before acclimation to the different pCO 2 levels, algae showed higher photosynthesis under high pCO 2 than adults from the northern site. However, after 30 days of exposure to high pCO 2 , adults from all sites showed a reduction in photosynthesis and growth. Nevertheless, adults from the northern site (higher upwelling) experienced a smaller decrease in growth in response to high pCO 2 . This can be explained by the fact that the tested high pCO 2 level is outside the natural range of environmental variability for the southern population. In contrast, populations at the northern site frequently experience pCO 2 levels higher than 1,100 μatm (Fig. 1). The differences in response of each population may be attributable to phenotypic plasticity (given the different environmental histories of the adults before collection) or to the effects of natural selection (Harley et al., 2012). Some have argued that within the natural range of pCO 2 variability, plasticity will play the major role in alleviating the effects of high pCO 2 , while outside the natural range, evolutionary and/or transgenerational effects may be more relevant (Calosi et al., 2013;Thor & Dupont, 2015). Only a few studies have tested for evolutionary adaptation to natural variation in pCO 2 (Kelly, Padilla-Gamiño, & Hofmann, 2013); however, the thermal tolerance literature abounds with examples of adaptive differences in thermal optima between populations, even when the tested temperatures do not fall outside the natural range of variability (Angilletta, Niewiarowski, & Navas, 2002;Sanford & Kelly, 2011). Therefore, it seems equally possible that the differences observed between populations in this study represent adaptive differences shaped by average differences in average pH, rather than the extremes. Using a broader range of partial pressures of pCO 2 will help to investigate whether there is a tipping point in the growth and calcification of Carpinteria Relative growth rate (%/day) High pCO 2 Low pCO 2 F I G U R E 5 Pigment content in adults of Corallina vancouveriensis exposed to high and low pCO 2 levels. Adults were collected from sites in California located north (Cambria and Arroyo Grande) and south (Santa Barbara and Carpinteria) of Point Conception (n = 12, mean ± SE). Solid bars represent treatments under high pCO 2 at the beginning of the experiment (black) and after 30 days (gray). Bars with dots and strips represent treatments under low pCO 2 at the beginning of the experiment (dots) and after 30 days (strips) in physiological decline (i.e., tipping point) in response to pCO 2 values above extreme values in the natural environment (Comeau, Edmunds, Spindel, & Carpenter, 2013). Moreover, future studies using a broader range of pCO 2 levels should use one header tank per experimental unit to achieve true replication . In our design, we had experimental units with interdependent treatment replicates (Design B4, , which may confound the effect of treatment with inherent differences between tanks. Reduced growth and/or calcification under high pCO 2 levels has also been observed in adults of the calcifying algal species C. pilulifera, Hydrolithon sp., Halimeda incrassata, Neogoniolithon sp., Lithotamnion corallioides, Arthrocardia corymbosa, and Porolithon onkodes (Anthony et al., 2008;Cornwall et al., 2013;Gao & Zheng, 2010;Gao et al., 1993;Johnson & Carpenter, 2012;Noisette et al., 2013;Ries, Cohen, & McCorkle, 2009;Semesi et al., 2009). However, elevated pCO 2 did not have an effect on growth in the intertidal coralline algae Corallina elongata (Egilsdottir et al., 2012), and in Litophyllum cabiochae, increased pCO 2 even enhanced calcification (Martin, Charnoz, et al., 2013;Martin, Cohu, et al., 2013). Lower growth rates in response to high pCO 2 may be a result of decreased calcification due to lower photosynthetic rates and higher respiration and/or increased dissolution associated with a lower saturation state. As the pCO 2 increases and saturation state decreases, it becomes more energetically costly to calcify. Our high pCO 2 treatments were nearly undersaturated with respect to calcite and aragonite, with Ω ara and Ω cal values close or at the saturation horizon (Ω = 1) (Kleypas, 1999). Differences in carbonate mineralogy could also explain the variation in growth responses among species; corallinales have high variability in mineralogical composition and can change their skeletal mineralogy in response to the local seawater chemistry (i.e., Mg 2+ concentrations) (Smith, Sutherland, Kregting, Farr, & Winter, 2012). As Mg-calcite is the most soluble form of CaCO 3 , higher Mg-calcite content in the skeleton can increase dissolution and reduce resistance to elevated pCO 2 (Morse et al., 2006).
Despite the fact that pCO 2 enrichment can stimulate photosynthesis by providing more substrate for carbon fixation, we observed lower photosynthetic rates (NPP and GPP) in response to high pCO 2 in C. vancouveriensis. This result is consistent with previous studies examining photosynthetic responses of coralline algae to high CO 2 levels (Anthony et al., 2008;Gao & Zheng, 2010;Gao et al., 1993;Martin, Charnoz, et al., 2013;Martin, Cohu, et al., 2013). Nevertheless, other studies have found enhanced photosynthesis under high pCO 2 (Semesi et al., 2009) or no effects at all (Egilsdottir et al., 2012;Hofmann et al., 2012;Semesi et al., 2009). Photosynthesis in the Corallina genus is not strictly carbon-limited; they have evolved processes such as the CCMs that allow them to increase the amount of CO 2 around Rubisco (enzyme involved in fixing CO 2 during photosynthesis) via ion channels or by catalyzing the transformation of HCO − 3 into CO 2 (Giordano et al., 2005;Raven, Beardall, & Giordano, 2014;Raven et al., 2012).
Differences within and between algal species in response to high pCO 2 may be due to the presence and activity of these CCM (Giordano et al., 2005) and whether they involve external or intracellular carbonic anhydrase (Reinfelder, 2011); an enzyme that catalyzes the interconversion of dissolved bicarbonates and carbon dioxide. Furthermore, it has been shown that in coralline algae, the primary carbon used in photosynthesis is HCO − 3 (Borowitzka, 1981), and in C. pilulifera, calcification and photosynthesis can be enhanced by carbonate (CO 2− 3 ) and bicarbonate (HCO − 3 ) but not by the addition of free CO 2 (Gao et al., 1993). In the coralline algae, C. officinalis carbonic anhydrase (CA) activity was ~40% higher in individuals grown under high pCO 2 than individuals grown in ambient conditions for 4 weeks (Hofmann et al. 2012) contrary to the expectation that CA would be downregulated when more pCO 2 was available. However, after a long-term exposure (12 weeks), C. officinalis showed an inverse trend between CA activity and pCO 2 concentration. Future studies are needed to better understand how CO 2 can be regulated and concentrated at the Rubisco fixation site and how elevated CO 2 will impact the activity of CA and the operation and interactions of CCM (Koch et al., 2013).
Respiration rates of C. vancouveriensis from all tested populations increased after culturing them under high pCO 2 for 30 days, indicating greater physiological demands for algae growing under this treatment. Our results are consistent with Noisette et al. (2013), who found increased respiration, lower calcification, and higher occurrence of bleaching in response to high pCO 2 in the intertidal coralline alga Lithophyllum incrustans. These studies suggest that high pCO 2 increases metabolic demands and that poor physiological state could affect the calcification balance and increase susceptibility to other stressors (Martin, Charnoz, et al., 2013;Martin, Cohu, et al., 2013;Noisette et al., 2013), possibly compromising long-term survival and reproduction. Conversely, other studies on coralline algae did not find effects of high pCO 2 on respiration (Egilsdottir et al., 2012;Hofmann et al., 2012;Johnson, Moriarty, & Carpenter, 2014;Martin, Charnoz, et al., 2013;Martin, Cohu, et al., 2013;Noisette et al., 2013;Semesi et al., 2009). Higher respiratory rates under high pCO 2 in our study were associated with lower net and gross photosynthetic rates and lower chlorophyll and phycobiliprotein content. Chl a, phycocyanin, and phycoerythrin decreased ~8%, 14% and 11%, respectively, under high pCO 2 , indicating a reduction in photosynthetic potential (capacity to absorb light) for the four populations studied after the 30-day acclimation. Interestingly, L. incrustans did not show differences in photosynthesis or chlorophyll content between pCO 2 levels (Noisette et al., 2013). Carotenoids, which also absorb light energy for photosynthesis and protect the chlorophyll from photodamage, were not affected by high pCO 2 . Similar results for carotenoids under high pCO 2 were found in C. sessilis and the diatom Phaeodactylum tricornutum (Gao & Zheng, 2010;Li, Gao, Villafañe, & Helbling, 2012).
To understand how physiological responses can impact individual and population responses to ocean acidification, we also need to consider the different thresholds and limits of different life stages, not just the adults. Currently, little is known about the effects of ocean acidification on early life-history stages of coralline algae, despite the fact that these stages can be very susceptible to high levels of pCO 2 (Bradassi et al., 2013;Cumani et al., 2010;Kroeker et al., 2012;Kuffner et al., 2007;Roleda et al., 2015) and possibly act as demographic bottleneck for benthic recruitment under acidified conditions. In California, adults of C. vancouveriensis release spores throughout the year with no seasonal trends (Miklasz, 2012). Spores of coralline algae are not calcified; they are negatively buoyant and attach to the substratum using developing filaments that attach to surface microstructures (Steneck, 1986), or using mucilage and epoxy-like resins that can harden over time (Fletcher & Callow, 1992). Spores of C. vancouveriensis can fully attach after 24 hrs of settlement, but some spores can achieve attachment within 1 hr of release (Miklasz, 2012). Once they have attached in the substratum, they flatten, and calcification starts after the first cell division (Walker and Moss 1984).
Germination can occur in as little as 8 hr (Miklasz, 2012) suggesting that C. vancouveriensis has limited dispersal and a high potential for local adaptation (Hereford, 2009;Leimu & Fischer, 2008).
Our results show that early life-history stages of C. vancouveriensis were more resilient to the direct effects of near-future acidification levels than adults. However, future work will need to be performed to eliminate the possibility that these differences are due to seasonality, because experiments with adults were performed around 2 months earlier than experiments with spores.
Spore growth and survival did not differ between pCO 2 treatments; however, spores at each site had different growth rates.
Interestingly, growth rates did not differ consistently between regions; the largest and smallest growth rates were found in algae from sites at the northern region (150% and 50% growth, Cambria and Arroyo Grande, respectively), whereas algae from the southern region had more similar growth rates (116% and 97% growth, Santa Barbara and Carpinteria, respectively). Differences in growth between sites may be due to maternal effects or variability in growth requirements between sites. Likewise, Bradassi et al. (2013) found no differences in growth in early life stages of the CCA Phymatolithon lenormandii, but observed increased mortality, abnormal thalli, and different calcification patterns (margin of the thallus vs. total area of the thallus) under high pCO 2 levels. Negative impacts of high pCO 2 were also seen in early life stages of the crustose coralline L. incrustans including low spore production and growth and increased mortality of the germination disks (Cumani et al., 2010). Furthermore, in recruits of A. corymbosa (articulate coralline algae), lower growth rates and decrease in Mg-calcite content were observed under low pH treatments (Roleda et al., 2015). However, these responses were not as pronounced as in the adults of the same species (Cornwall et al., 2013) suggesting that juveniles may be more resistant than adults to lower pH. In mesocosm experiments, Kuffner et al. (2007) found that recruitment of tropical crustose coralline algae decreased 78% with elevated seawater carbon dioxide concentration, whereas Kroeker et al. (2012) found no effects of low pH in the recruitment of temperate coralline algae settling in plates located at a volcanic CO 2 site. These contrasting results highlight the variability in sensitivity of early stages of different species in response to elevated pCO 2 and the limitations to projecting individual and population-level responses to ocean acidification without considering variation in tolerances of different life-history stages within a species (Harley et al., 2012). However, Kuffner et al. (2007) used diluted hydrochloric acid to reduce pH of the water in the experimental treatments, which could lead to different outcomes than if CO 2 was used to acidify the water. It is also important to note that differences in physiological response between populations may not only be attributed to differences in pCO 2 natural conditions between sites but also to the simultaneous exposure of other parameters associated with upwelling events such as low temperatures and high nutrient concentrations. Future experimental studies manipulating pCO 2 to simulate long-term environmental variability should be performed to conclusively differentiate the effects of pCO 2 between populations.
Our work highlights the importance of considering complete life cycles when projecting the biological impacts of future environmental changes, because different stages will have different physiological thresholds and tolerance limits. Our study suggests that spores are less sensitive to high pCO 2 than adults of C. vancouveriensis. The physiology and growth rates of adults were impacted at the highest pCO 2 . The tested scenarios were only relevant in the context of ocean acidification for the southern sites experiencing lower upwelling, and individuals from these populations were more sensitive to the high pCO 2 . Adults from the northern populations are already experiencing high pCO 2 tested in this study and as predicted were more tolerant to these levels.