Contrasting responses of photosynthesis and photochemical efficiency to ocean acidification under different light environments in a calcifying alga

Ocean acidification (OA) is predicted to enhance photosynthesis in many marine taxa. However, photophysiology has multiple components that OA may affect differently, especially under different light environments, with potentially contrasting consequences for photosynthetic performance. Furthermore, because photosynthesis affects energetic budgets and internal acid-base dynamics, changes in it due to OA or light could mediate the sensitivity of other biological processes to OA (e.g. respiration and calcification). To better understand these effects, we conducted experiments on Porolithon onkodes, a common crustose coralline alga in Pacific coral reefs, crossing pCO2 and light treatments. Results indicate OA inhibited some aspects of photophysiology (maximum photochemical efficiency), facilitated others (α, the responsiveness of photosynthesis to sub-saturating light), and had no effect on others (maximum gross photosynthesis), with the first two effects depending on treatment light level. Light also exacerbated the increase in dark-adapted respiration under OA, but did not alter the decline in calcification. Light-adapted respiration did not respond to OA, potentially due to indirect effects of photosynthesis. Combined, results indicate OA will interact with light to alter energetic budgets and potentially resource allocation among photosynthetic processes in P. onkodes, likely shifting its light tolerance, and constraining it to a narrower range of light environments.

calcification was calculated for each sample from its change in buoyant weight, normalized to the number of days it was exposed to experimental treatments, as well as its surface area. (As in 2014, the density of calcite was used to estimate the change in skeletal dry weight of each sample). Net photosynthesis and light-adapted (post-illumination) respiration were measured over the course of 4 days for 9 CCA from each pCO2 x light treatment, starting after the samples had been in their experimental treatments for two weeks. Photochemical efficiency and P-E curve measurements were not made in this pilot experiment.

Results
Light had a significant effect (p < 0.001) on gross photosynthesis (calculated by adding the rate of oxygen consumed by respiration to the rate of oxygen evolved by net photosynthesis), causing a 46% increase in photosynthesis between the low and high light treatments. However, pCO2 did not affect photosynthesis, nor did the interaction between pCO2 and light (Fig. S1). In contrast, light-adapted respiration showed limited variability, increasing on average by 13% between the low and high light treatments. However, it was not significantly influenced by light, pCO2, or the interaction of these factors (Fig S2) .
Net calcification declined by 18% in the elevated pCO2 treatment, leading to a significant difference between the ambient and elevated treatments (p = 0.015). There was also a significant effect of light on calcification (p = 0.009), leading to a 26% increase in calcification between the low and high light treatments (Fig. S3). However, there was no significant light x pCO2 interaction. (See Table S2 for a summary of statistical results.)

Discussion
Overall, despite differences in pCO2 and light treatments, the results of the 2013 pilot study were qualitatively similar to the 2014 experiment. Like in 2014, light-adapted respiration and gross photosynthesis did not respond to pCO2 (Fig. S1, S2), but net calcification was lower in the elevated pCO2 treatment (Fig. S3). This decline in calcification was similar in magnitude across experiments (15-18%). Additionally, in both experiments light generally had a larger effect than pCO2, leading to a 22-26% increase in calcification, a 27-46% increase in photosynthesis, and a 13-43% increase in respiration between low and high light treatments.
However, despite the positive correlation between light and photosynthesis and calcification, treatment light level did not influence the response of any of the measured physiological rates (photosynthesis, respiration, calcification) to pCO2 (Table S2). Thus, like the 2014 experiment, light level did not modify the sensitivity of photosynthesis, respiration, or calcification to OA.

Differences between the two experiments
However, despite similarities, there were some differences between the 2013 and 2014 experiments. First, respiration did not increase significantly with light in 2013 (Table S2) vs. n = 16), leading to lower power to detect a significant effect of light.
Additionally, the rates of respiration, photosynthesis, and calcification were higher in 2013 compared to 2014; (rates decreased on average by 5.5% in 2014 for calcification, and by 50% for respiration and photosynthesis). Several factors could explain this variation, including differences between the experiments in: 1) experimental duration, 2) tank water turnover rates, and 3) water quality. Exposure to experimental treatments was longer in 2014, potentially reducing growth and metabolic rates as the CCA acclimated to their new experimental conditions, which had lower light, flow, etc. than their natural habitat. Supporting this hypothesis, gross photosynthesis rates measured for the P-E curves in 2014 were generally higher (by ~ 0.1 µmol O2 cm -2 hr -1 for the ambient pCO2 treatment) than rates measured at similar light levels 1-2 weeks later. Additionally, water turnover rates were higher in the experimental tanks in 2013, which also may have facilitated faster growth and metabolism in the CCA that year by increasing nutrient delivery, gas exchange, and metabolite removal. Finally, differences in water quality between the years (nutrients, pollution, etc.) that we couldn't control for with our water filtering procedure, also could have contributed to differences in physiological rates. However, no direct evidence suggests this was the case, so the first two explanations seem the most likely. Despite these differences between years, the rates we measured in both experiments were within the ranges reported for P. onkodes [1][2][3][4][5] , as well as other CCA 6 .
Overall, our results indicate that the effects of OA on gross photosynthesis, light-adapted respiration, and net calcification in P. onkodes are consistent across a range of experimental conditions, but are not necessarily larger in magnitude than the effects of other factors such as flow or light .   Tables for the 2013 pilot study   Table S1) Physical conditions of the 2013 pilot study. Values represent group means ± SEM. There were 4 tanks per pCO2 treatment, and light treatments were nested within tanks. Treatments abbreviations are as follows-ambient pCO2 (ACO2) or high (HCO2); low light (LL), medium light (ML), or light saturated, i.e., "high" light (HL). Physical variables include total alkalinity (TA), saturation state of seawater with respect to calcite (Ω calcite), and photon flux density (PFD) of light treatments.

CCA identification and characterization of in situ light levels
Methods S1 -CCA species identification P. onkodes crusts were identified in the field using morphological characteristics such as its thick, heavily calcified crust that strongly adhered to the substrate, its smooth, chalky epithallial texture, tight trichocyte fields, and very small to non-obvious conceptacles 7,8 .
Specimens were collected to confirm identifications in the lab. Species identifications were done under the supervision of Dr. Robert S. Steneck, by analyzing specimens under a dissecting microscope. Features such as conceptacle size and shape, distribution of trichocytes, and cell shape, size, and structural arrangement (observed in cross section), allowed us to confirm the identification of the specimens as P. onkodes.
Since conducting this study, other researchers have determined that P. onkodes is a complex of greater than twenty morphologically similar species, rather than a single pantropically distributed species 9 . Since we did not preserve samples for genetic analysis, we cannot confirm the precise species identity of our CCA. However, all samples were collected from similar microhabitats (open horizontal and vertical reef substrates) at a single back reef location in Moorea, French Polynesia, suggesting they were likely a single (or very few) species.
Therefore, the results of our study should be interpreted as applying to an undefined P.
"onkodes" species, or a complex of several morphologically similar species.

Methods S2 -In situ light levels
Light levels characterizing the natural habitat of the experimental CCA were measured on five separate sunny days, within an hour of local apparent noon at the experimental collection site. Photon flux densities were measured at the location of P. onkodes crusts (or scars where crusts were recently collected), using a 2π PAR sensor attached to a diving-PAM (Walz, Germany), held parallel to the surface of the CCA. After recording the light level at a crust, the researcher swam at least 1 m in a haphazard direction until another P. onkodes crust was found.
Light measurements were made only for P. onkodes in exposed reef habitats (including horizontal and vertical substrates), to remain consistent with the environment where the experimental CCA were collected. This procedure excluded some individuals that were observed in semi-cryptic microhabitats, e.g. under coral overhangs or in crevices in the reef substrate. Therefore, these measurements provide a snapshot of the light levels that the experimental CCA experienced in their natural environment, but do not characterize the full range of light environments inhabited by P. onkodes in the Moorean back reef.

Methods S3 -Respirometry
Oxygen concentration was automatically corrected for temperature using a PreSens temperature probe, and adjusted for the average salinity of the experiment. The oxygen probe was calibrated daily using a two-point calibration of 0% and 100% oxygen, using zero-oxygen seawater and water-saturated air, respectively. Zero-oxygen seawater was produced by supersaturating seawater with sodium dithionite (Na2S2O4). Water in the chambers was mixed at a constant rate using rotating stir bars, and was kept at an approximately constant temperature using an external circulating water bath. Samples were acclimated to the chambers for 10 minutes before their metabolic rates were measured. Measurements of photosynthesis and lightadapted respiration were made over the course of 5 days at the end of the experiment between 07:00 and 19:30 hr. Dark-adapted respiration was measured as part of the P-E curve incubations, and occurred 2 -7 days before net photosynthesis and light-adapted respiration. A stratified random design was used to determine the order in which samples from different treatments were run, in order to eliminate the influence of incubation timing on the measured metabolic rates.

Methods S4 -P-E curves
P-E curve measurements began after the samples had been in their experimental treatments for two weeks, and then were made over the course of 6 days between the hours of 06:45 and 20:45 hr. As with the previously described respiration and photosynthesis measurements, a stratified random design was used to determine the order that samples from different treatments were run, to eliminate the influence of incubation timing on P-E curve characteristics. For each algal sample, respiration rates were measured first, after the alga had been dark-acclimated for at least 45 minutes to eliminate the stimulatory effect of photosynthesis on respiration 10 . Following respiration, incubations were conducted on the same sample at each successively increasing PFDs, allowing the sample to acclimate to the new light level for at least 10 minutes before new measurements began. For each sample, seawater in the incubation chamber was completely replaced with fresh seawater from the treatment tank after every third incubation, or after the sample had spent longer than 105 minutes in the chamber, whichever occurred first. Due to the slow metabolic rates of CCA, and based on the PreSens logger readings of oxygen concentrations, this frequency of water replacement was deemed adequate to prevent changes in the chemical environment of the chambers of large enough magnitude to influence the physiological rates of the CCA.

Methods S5 -Photochemical Efficiency (Fv / Fm)
Because of the large number of samples (n = 192), these measurements were taken over the course of four nights at the end of the experiment. Each night, algae were dark-adapted for two hours following the onset of the dark period, and then for a randomly chosen subset of samples from each light treatment in each tank, measurements were taken from a flat area in the center of each algal thallus. These measurements were made with the 5.