Seasonal Preservation Success of the Marine Dinoflagellate Coral Symbiont, Symbiodinium sp.

Coral reefs are some of the most diverse and productive ecosystems on the planet, but are threatened by global and local stressors, mandating the need for incorporating ex situ conservation practices. One approach that is highly protective is the development of genome resource banks that preserve the species and its genetic diversity. A critical component of the reef are the endosymbiotic algae, Symbiodinium sp., living within most coral that transfer energy-rich sugars to their hosts. Although Symbiodinium are maintained alive in culture collections around the world, the cryopreservation of these algae to prevent loss and genetic drift is not well-defined. This study examined the quantum yield physiology and freezing protocols that resulted in survival of Symbiodinium at 24 h post-thawing. Only the ultra-rapid procedure called vitrification resulted in success whereas conventional slow freezing protocols did not. We determined that success also depended on using a thin film of agar with embedded Symbiodinium on Cryotops, a process that yielded a post-thaw viability of >50% in extracted and vitrified Symbiodinium from Fungia scutaria, Pocillopora damicornis and Porites compressa. Additionally, there also was a seasonal influence on vitrification success as the best post-thaw survival of F. scutaria occurred in winter and spring compared to summer and fall (P < 0.05). These findings lay the foundation for developing a viable genome resource bank for the world’s Symbiodinium that, in turn, will not only protect this critical element of coral functionality but serve as a resource for understanding the complexities of symbiosis, support selective breeding experiments to develop more thermally resilient strains of coral, and provide a ‘gold-standard’ genomics collection, allowing for full genomic sequencing of unique Symbiodinium strains.


Introduction
Coral reefs, some of the oldest and most diverse ecosystems on our planet, are under siege. Climate change is contributing to bleaching and acidification, that leads to stress and degradation, Our pilot work for the present study supported this view. Specifically, Symbiodinium extracted from two Hawaiian species, Fungia scutaria and Porites compressa, were cryopreserved using standard slow-freezing rates of 0.1, 1, 3, 20 or 100°C/min, all of which resulted in no post-thaw viability (defined as having an intact photosystem after thawing as measured with a pulse amplitude fluorometer). One of the more important uses of cryopreserved Symbiodinium might be using different clades to help the coral survive changing conditions. Toward that goal, 4-day old F. scutaria coral larvae initially took up native extracted, cryopreserved and thawed native Symbiodinium type C1f, but 2 days later had eaten or rejected these cells, suggesting there was something wrong with them.
The combined observations that these Symbiodinium species appeared chill-and cryoprotectant sensitive as well as rejected by the coral host upon thawing clearly indicated the need for a more novel preservation approach. Vitrification, a non-equilibrium, ultra-rapid cooling technique, appeared prudent to examine as this procedure 'outruns' the chilling sensitivity element for most cells while often providing improved survival for many cell types as well as embryos [38]. The major factors involved in vitrification are: 1) the stepwise addition of highly concentrated solutions of cryoprotectants and sugars that dehydrate the target cell; and 2) ultra-rapid cooling of the cell suspension (e.g., 1,000°C/min) to form a transparent glass-state rather than ice crystals [38]. Over the past 30 years, traditional methods of vitrification have transformed assisted reproductive techniques for handling human embryos which are now vitrified with a post-thawing success of >90% [39]. Until recently, human eggs were not well preserved by traditional vitrification processes because of a high lipid content imposing extreme sensitivity to chilling temperatures [40]. Wedding classical vitrification protocols with the ultra-fast thawing rates (e.g., 10,000°C/min) of new films, called Cryotops has now made human oocyte cryopreservation highly successful [40]. Like human oocytes, Symbiodinium contain an abundance of lipids [41] that may be contributing to their chilling vulnerability [37]. Thus, vitrification may help with cryopreserving Symbiodinium, especially as new cryococktails are now available that form excellent glassy states that, in turn, may reduce toxicity and osmotic stress.
Our goal was to elucidate the biological mechanisms required to successfully cryopreserve Symbiodinium extracted from three Hawaiian coral species, specifically F. scutaria, P. compressa and Pocillopora damicornis. Symbiodinium from F. scutaria was used as a model for triaging a suite of candidate cryoprotectants and vitrification protocols that were maintained post-thaw over 90 min. Based on those results, we developed a set of vitrification protocols that allowed Symbiodinium to survive for at least 24 h post-thaw. Then, the best candidate protocol was tested throughout the year to determine a potential role of seasonal variation on the success of the Symbiodinium to withstand freezing. Additionally, the candidate protocol was tested side-by-side with accepted slow-freezing methods [35] for Symbiodinium to compare their post-thaw results at 24 h. And finally, to determine the applicability of our process across species, we conducted a comparative vitrification study exposing the Symbiodinium from F. scutaria, P. compressa and P. damicornis to our candidate vitrification protocol, assessing their 24 h post-thaw viability.

Coral Collection and Husbandry
Whole individuals of F. scutaria, colonies of P. damicornis and fragments of P. compressa were collected from the shallow reef flats around Coconut Island in Kaneohe Bay, Hawaii from 2011 to 2015. Different locations on the reef flat were chosen for collection to ensure as much coral genetic diversity as possible. Coral were collected from the Bay and used immediately or maintained in flowing seawater tables connected directly to Kaneohe Bay with natural temperature and light exposure throughout the study. Collection was performed with the appropriate permits from the state of Hawaii's Department of Land and Natural Resources (Special Activity Permit # SAP 2011: 2011 -1, 2012: 2012-63, 2013: 2013-47 and 2014: 2015-17). No institutional ethical approval was required for any of the experimental research described herein.

Monitoring and Preparing Symbiodinium Samples
Symbiodinium were extracted from live coral tissue, cleaned and their health monitored according to the methods of Hagedorn et al 2010 [37]. Briefly, a Junior Pulse Amplitude Modulated fluorometer (Junior-PAM, Walz, Germany) provided an indication of the viability of photosynthesis in Photosystems I and II. Settings on the Junior-PAM, specifically the gain and intensity settings, were kept constant throughout each and across all experiments in these studies to ensure that we were observing physiological changes. The percent change from a sample's initial quantum yield post-treatment was reported. For the purpose of this paper, any quantum yield or Y-value at or below 0.050 was defined as non-viable and as having a non-functional photosystem [37]. By inference, quantum yields with the least change from their initial measures, indicate a robust and healthy Symbiodinium sample, whereas those close to 100% change were considered dead. Measurements with the Junior-PAM were conducted under consistent fluorescent lighting in the laboratory at mean photosynthetically active radiation (PAR) value of 4.6 ± 0.51 SEM μmols/m 2 /s (n = 5, Apogee Instruments, Model MQ-200). All solutions, including those containing the extracted Symbiodinium, were made in and maintained in 0.2 μmfiltered seawater (FSW), unless otherwise stated.
Cryotop (Kitazato Corporation, Tokyo, Japan) methods have been used to manipulate human oocytes for ultra-rapid warming to produce successful vitrification [40]. To determine the applicability of these devices to cryopreserving Symbiodinium, we used a 2-μl drop of zooxanthellae on the Cryotop (Experiment 1 only) or a 1.5 μl drop spread as a thin agar film the following four treatments with Symbiodinium embedded within as mentioned in Fig 1 (used in all other Experiments) melted on to the top of the Cryotop. Conditions (i.e. temperatures and times for embedding) are given below in part 2. With this agar-and-Symbiodinium film method, the Symbiodinium did not fall off during the passage from solution-to-solution and during freezing or culture because they were embedded within the agar, making Symbiodinium manipulations easier. The conditions (i.e. temperatures and times for embedding) are given below in part 2. The agar-and-Symbiodinium film covered an area of the Cryotop approximately 0.5 cm in length from the tip of the Cryotop surface and had a domed shape (see schematic in Fig 1). During these protocols, Symbiodinium health was assessed after loading into agar and onto a Cryotops and then again during certain intervals post-thaw. Cryotops samples were moved through solutions and afterwards set into shallow dishes to await viability assessment with the PAM.

Experiments
Experiment 1: Vitrification of Symbiodinium in a spherical pellet. Our previous studies [37] demonstrated that a 10% dimethyl sulfoxide solution (a penetrating cryoprotectant) appeared to minimize toxicity for Symbiodinium, while requiring about 60 min to equilibrate within the cell. To increase osmolality of the cryoprotectant, 0.5 M glucose or trehalose (as a non-penetrating cryoprotectant that helped dehydrate the cell by removing intracellular water) was added to 10% V/V dimethyl sulfoxide (to help stabilize and protect membranes). Therefore, all subsequent trials in this experiment used dimethyl sulfoxide (to help stabilize and protect membranes) often in tandem with 0.5 M glucose or trehalose.
Symbiodinium were extracted and cleaned [37], counted with a hemocytometer, and suspended in the following four treatments at a concentration equaling~5 x 10 7 cells/ml. The Symbiodinium were partitioned into four treatment subsamples and their initial quantum yield measured. These treatments were (1) FSW, (2) FSW + freezing, (3) CPA = cryoprotectant exposure, no freezing, and (4) Vitrification = cryoprotectant exposure + freezing. Six different vitrification protocols (defined below) were tested as the treatment 4. In all experiments the solutions and cryoprotectants used for treatment 3 and 4 were the same, except treatment 3 was not vitrified. All of the vitrification protocols had the same first step, which was a 10% dimethyl sulfoxide solution for 1 h followed by centrifugation (Eppendorf Centrifuge 5415D at 5900 x g for 5 min). Then the cells were resuspended in dehydrating step 2 solutions which differed and were either: 1) 10% dimethyl sulfoxide (5 min exposure); 2) 10% dimethyl sulfoxide + 0.5 M trehalose (5 min exposure), 3) 10% dimethyl sulfoxide + 1 M trehalose (5 min exposure), or 4 to 6) 10% dimethyl sulfoxide + 0.5 M trehalose + 0.5 M glucose (for 5, 10 or 15 min exposures). After step 2, a 2 μl drop of Symbiodinium and solute containing 1 x10 5 cells was placed onto a Cryotop and immediately immersed in liquid nitrogen until the boiling stopped (Fig 1, upper schematic). Once frozen, the cells and Cryotops were immediately moved to 98 μl of FSW (23 to 25°C) in a 1.5 ml plastic tube. During freezing and warming, the Symbiodinium drop situated on the Cryotop was observed visually a few cm below the surface of all liquids for the presence of ice crystals (ice crystals form obvious white patterns viewable with the naked eye). Cells were maintained for 24 h in FSW. All samples were loosely covered over night to reduce stress from overt light input. However, the cover was typically removed about 15 min before the final assessment with the Junior-PAM. These samples were not dark-adapted. Once a sample's Y-value was measured after a treatment, a percent change from its initial value was calculated.
Experiment 2: Vitrification of Symbiodinium in a laminar agar film. The design of Experiment 1 had three challenges. First, the Symbiodinium had to be centrifuged to transfer the cells from solution to solution potentially increasing cellular stress. Second, when the cells were thawed, the vitrification solution was only diluted 1:50, potentially adding to osmotic stress during the 24 h maintenance period. Third, the Symbiodinium were in a spherical profile, which increased the surface to volume ratio, potentially reducing permeation. To address these issues, we changed the Symbiodinium holding conditions to a laminar profile, thus reducing the surface-to-volume ratio. Specifically, we mixed Symbiodinium 1:1 in agar at different agar concentrations (1.0, 1.25, and 1.5%) and at various starting mixing temperatures (45, 55, and 65°C). We tested 24 h survival to ensure that the Symbiodinium's photosystem remained intact and that they stayed on the Cryotop during all vitrification manipulations at a final concentration in the agar of 1 x 10 7 cells/ml (n = 32, F. scutaria). The sample's quantum yield was measured initially and then again after an agar-Symbiodinium Cryotop was exposed to its treatment solution and maintained for 24 h in FSW. A percent change from its initial value was calculated. As above, the samples were loosely covered over-night and assessed 24 h later.
To understand how the change in the quantum yield related to the percentage of Symbiodinium left alive in the sample, a live: dead curve was created, as in a previous study [37] that demonstrated a linear relationship between the percent live cells in the sample and normalized quantum yield. However the earlier samples were measured in suspension and not in a laminar film, and we did not know if this might have an affect on the PAM readings. Therefore, a parallel control experiment was conducted to understand how the quantum yield values related to viability of the Symbiodinium within the physical measurement parameters on the Cryotop (i.e., Cryotop agar films in 1 ml Eppendorf tubes with the Junior-PAM at a PAR of~5 μmols/ m 2 /s, 23 to 25°C). Symbiodinium from F. scutaria (n = 3) were extracted and cleaned [37]. Symbiodinium samples consisting of live: dead cells were created in a range of ratios from 1:0 (all alive) to 0:1 (all dead) and ratios in between (e.g., 10:1, 1:5, 2:3 and 4:1). Cells were killed by three cycles of microwave exposure followed by flash freezing and a final PAM assessment to ensure they were dead, and then the live and dead cells were mixed in known proportions. A starting sample (10 7 cells/ml) was split into two sub-samples. The first sub-sample was embedded in agar while the second was left in suspension. A sample's quantum yield was measured immediately and at 24 h. When the live:dead ratio was regressed against a normalized quantum yield, these two sub-samples produced the same linear results, as obseved previously [37] (P> 0.05, F = 0.0011, Linear Regression, see S1 Fig). Therefore, the physical parameters of the sample does not change the physiology of the Symbiodinium and generally a 50% change in the quantum yield values in these studies related to an approximately 50% reduction in viable cells in the sample.
Based on the results of Experiment 1, Symbiodinium extracted from F. scutaria (n = 4) were embedded in 1.5% agar on the Cryotop (as diagramed in Fig 1). As above, four treatments were assessed: (1) FSW, (2) FSW + freezing, (3) CPA = cryoprotectant exposure, no freezing, and (4) Vitrification = cryoprotectant exposure + freezing. Four different vitrification protocols were tested in treatment 4, as defined in Table 1. A candidate vitrification protocol was selected from these results bases on the percent change of the Symbiodinium's quantum yield and used in subsequent experiments. The rationale for using trehalose in these experiments it that this sugar is produced and exported by Symbiodinium [42], so it may naturally be playing some sort of protective role. Methanol was chosen because it is non-toxic to Symbiodinium [37].
In Culture Collections the post-thaw expansion of Symbiodinium in culture can take months, which may be due to the low number of cells surviving slow freezing cryopreservation. In this experiment, we compared the slow-freezing method [35] employed by many culture collections and our optimal vitrification process (above) to understand whether vitrification produced substantial numbers of viable cells post-thaw. Thus, Symbiodinium from F. scutaria (n = 2) were extracted and divided into two sub-samples and exposed to (1) our candidate vitrification protocol (Vitri-3, described above), and (2) the slow freezing method [35]. As before, the vitrification and slow freezing sub-samples were divided again into our standard four treatments: (1) FSW, (2) Frozen, no cryoprotectant exposure, (3) CPA = cryoprotectant exposure, no freezing, and (4) CPA plus slow freezing or vitrification + freezing. For the slow-freezing process, four 1 ml samples of Symbiodinium (1 x 10 7 cells/ml) were assessed initially with the Junior-PAM, centrifuged (Eppendorf Centrifuge 5415D at 900 x g, 5 min) to form a pellet, and the supernatant removed. Then, 1 ml of FSW was added to create treatments 1 and 2 (above) and 1 ml of 20% methanol was added to create treatments 3 and 4 (above). For treatments 2 and 4 designated for slow-freezing, the methanol was allowed to equilibrate into the cells for 10 min at 4°C. These samples were then placed into a -80°C freezer for 2 h, followed by immersion in liquid nitrogen for 10 min, bringing the samples down to -196°C. Each frozen sample was thawed by gentle agitation in room temperature water. At this stage, the cells were washed three times by concentrating them via centrifugation (900 x g, 5 min), removing the supernatant and supplementing them with1 ml of FSW. After the third wash, samples were covered with aluminum foil to block stray light and rested over night in FSW for at least 20 h. Then quantum yield values for each treatment were determined for the slow frozen and vitrified samples at the same time.
Experiment 3: Influence of season and the ability of Symbiodinium to survive vitrification. Findings from Experiment 2 identified the best candidate vitrification protocol for more detailed testing. As this experiment was being conducted during different times of the year, we had noticed variation in our quantum yields post-vitrification over time. We postulated that this variability might be due to seasonal changes in the physiological conditions of the Symbiodinium. Previous work on chilling sensitivity in coral fragments with intact symbionts [43] suggested that coral fragments were more tolerant to chilling in winter compared to spring [37], although this study was neither longitudinal nor conclusive. To determine the influence of season on the physiological response of Symbiodinium to vitrification, our candidate vitrification protocol (Vitri-3) was tested at least bi-monthly from February 2014 through January 2015. Specifically, we evaluated the change in the quantum yield after 24 h post-thaw from a minimum of five F. scutaria sampled per month throughout the year. Experiment 4. As Experiments 1 to 3 were conducted exclusively by extracting Symbiodinium isolated from F. scutaria, we decided to test the applicability of the vitrification to other species, specifically to other Symbiodinium (of Hawaiian coral origin) isolated from P. damicornis (n = 3) and P. compressa (n = 3) in parallel with our model F. scutaria (n = 3). Samples collected during the winter, the Symbiodinium samples were embedded in agar, incorporated onto the Cryotop and immersed in liquid nitrogen. As above, four treatments were assessed: (1) FSW, (2) FSW + freezing, (3) CPA = cryoprotectant exposure, no freezing, and (4) Vitrification = cryoprotectant exposure (using protocol Vitri-3 in Table 1) + freezing. The 24 h Vitrification of Symbiodinium maintenance in FSW post-thaw and viability analysis were the same as described in Experiment 2 and 3 above.

Statistics
All data analyses were performed using Prism 6.0 (Graphpad, San Diego, CA) and Excel (Microsoft, Redmond, WA). All percentage change data were log transformed prior to statistical analyses. For multiple group comparisons, normality was tested graphically and ANOVAs with a Tukey's, Bartletts or Dunnett's Multiple Comparison post-tests, a Kruskal-Wallis, or Linear Regression were performed (all post-tests were specified in the Results), and all data are expressed in mean ± SEM.

Experiment 1: Vitrification of Symbiodinium in a spherical pellet
None of the Symbiodinium exposed to the vitrification treatments described for this experiment survived; as there was no post-thaw quantum yield values > 0.05 after 24 h (Table 2). If the total osmolality of the solution was 3 M, the droplet failed to vitrify, potentially causing lethal intracellular ice damage. An increase in solution osmolality to > 3 M did not improve post-thaw quantum yield following Symbiodinium vitrification. Even extending the dehydration time (during Solute Step 2) failed to increase post-thaw viability (Table 2). Therefore, although the sample drop appeared grossly vitrified, there was no post-thaw viability, suggesting intracellular damage. The data also appeared to suggest that most treatments were imposing little toxicity in the Symbiodinium, because all solution exposures (CPA column, Table 2) had only modest changes in quantum yield values (range, 0 to 30%). As there also was no clear preference for either glucose or trehalose (  vitrification (S2 Fig). Agar heated to 65°C adhered best and had the fewest failed treatments (6.3, 9.4, and 21.9% failed treatments for 65, 55, and 45°C initial agar heating, respectively). Therefore, the 1.5% agar solution heated to 65°C was chosen for all subsequent experiments. When Vitri-1 to Vitri-4 solutions were compared, only the Vitri-3 protocol produced viability results similar to FSW alone or CPA and vitrification solution but no freezing (Fig 2; P< 0.05, ANOVA, F = 21.63, Dunnett's Multiple Comparison Test). Therefore, the Vitri-3 protocol was used in all further experiments. Moreover, using the same Symbiodinium samples only vitrification with Vitri-3 produced viable cells 24 h post-thawing, whereas slow freezing method [35] did not (S3 Fig).  (Fig 3). On average, there was a 37 and 44% change in quantum yield in the winter and spring, respectively, compared to 60% for both the summer and fall (P> 0.05, ANOVA, F-= 5.16; Tukey's Multiple Comparison Test, Fig  3A). The environmental conditions that might influence seasonality in the Symbiodinium are PAR light level and water temperature at Coconut Island (Fig 3B and 3C).

Experiment 4
The Vitri-3 protocol using the Cryotops produced survival in the Symbiodinium of all three species tested. Survival in P. damicornis was higher (P < 0.05, Kruskal-Wallis test, Fig 4) with a  mean change in quantum yield of 18.6%, compared to 56% and 47% for F. scutaria and P. compressa, respectively (P > 0.05, Kruskal-Wallis).

Discussion
This study was designed to explore the efficacy of vitrification for the cryopreservation of the marine dinoflagellate coral symbiont, Symbiodinium sp.. These algal symbionts are essential for maintaining the health and viability of most warm water coral around the globe. During this study, we made several significant discoveries including: (1) the creation of a method to safely and quickly cryopreserve Symbiodinium by ultra-rapid freezing or vitrification; (2) the means to quickly assess the viability and estimate the number of viable algal cells 24 h post-thaw, using PAM fluorometry; (3) the finding that the success of this preservation method had a seasonal physiological component; and, (4) the successful efficacy of vitrification was applicable to Symbiodinium from multiple coral species. Thus, this protocol appeared capable of safely preserving the Symbiodinium of multiple coral species and, when collected during winter, resulted in a post-thaw survival of at least 50 to 90%, in the summer and winter, respectively. Although a slow-freezing cryopreservation method exists for Symbiodinium [35], we postulate that few cells survive the slow-freezing process and may be the reason it takes months to culture them post-thaw by this method. In our comparison study (S3 Fig), vitrified samples produced a 58% change in quantum yield compared to 99% for slow freezing, indicating that the latter cell population was largely dead. If we compared the resulting numbers, and assumed that (1)~40% of the cells were alive in the vitrified samples at 24 h and (2) the volume of the agar on the Cryotop was~1.5 μl with an initial concentration of 10 7 cells/ml, then we estimated that~6.0 x 10 3 were alive in the vitrified samples post-thawing and at 24 h. This is a large number to begin the culture process. By contrast, the number of cells alive in the slow frozen samples was below detection via PAM, so probably few cells remained alive.
The seasonal variation in the success of the vitrification process for Symbiodinium suggested that sampling and vitrification of Symbiodinium should occur ideally in winter. Other types of seasonal variation in physiological responses has been observed in lipid-rich microalgae grown outside [45]. Specifically, these microalgae show distinct shifts in their total lipids, found in the membrane and in vacuole, which were 11% in winter and 30% in fall, explained mostly by light and temperature [45]. We observed this same seasonal pattern of light and temperature correlating with the vitrification success in Symbiodinium. This seasonality is not surprising given that the amount of sunlight stimulating the symbiont's photosystem is less in winter, potentially altering membranes or fat synthesis. In some animals, both the physiology and preservability of their germplasm can be seasonal [46,47], as well. Additionally, cold hardiness produces significant changes to the extra-and intracellular physiology of somatic cells of plants [48] and animals [49,50], allowing them to endure extreme cold during the winter months. For example, the mechanisms underlying cold-hardiness are now beginning to be understood for the frog [50], because as winter approached two cryoprotective agents, urea and glucose, are synthesized and mobilized to the cells to protect them from the chilling temperatures. Taken together, is seems reasonable that the Symbiodinium may produce more intracellular lipids during the high-light intensity and warm summer and fall months, thus increasing their chilling sensitivity and reducing their overall vitrification success. However, these physiological changes in lipid content remain to be tested to verify this mechanism.
Cryobanks reflect a new and major type of preservation that can be added to conventional museum archives, but in this case, the living biomaterials go beyond dried materials to include gametes, embryos, somatic cells, blood, and DNA. Most species' genomic material, especially that of marine species, needs to be captured in the field far from sophisticated equipment. Vitrification is a field-friendly process that does not require a lot of equipment. Because these cells cannot often be captured and brought into the laboratory alive, cell preservation methods, such as cryopreservation, are paramount. During more standard preservation processes, such as immersion in alcohol, as is common for cells or tissues currently stored as museum specimens, DNA can often break into pieces over time [51]. This makes an extraordinarily strong and urgent case for creating cryobanks for cells, where they remain, frozen, but alive with near perfect genomic constructs. However, the future proof of concept for a global vitirified Symbioninium bank will be whether these cells expand in culture and are absorbed and maintained by coral larvae post-thaw. Both studies are currently ongoing.
Decisive conservation actions are needed to save reefs, with the first priority being habitat preservation. However, corals now face global rather than only local threats, requiring futurethinking tools, including ex situ conservation practices that can protect extant species and genetic diversity and integrity. In this context, we (and others) have argued for the biological banking of coral and their symbionts [52,53]. This is critical in light of the ever increasing numbers of bleaching events since 1980 [3]. To date, only coral sperm can be cryopreserved [53] because of the large amounts of lipids in coral eggs and embryos, leading to chilling sensitivity inhibiting viability, but the ultra-rapid freezing of Cryotops, using cells in a more laminar distribution on the surface, hold great promise for out-running this sensitivity potentially allowing success in the future. A cryobank not only would serve to protect viable specimens of each high priority species, but also could be used to provide biomaterials to scientists for both basic and applied research. Meanwhile a portion of the bank could retain frozen but alive specimens for hundreds of years, if necessary, in liquid nitrogen, allowing the time necessary to mitigate threats and restore habitats. We also see thawed samples from the repository being used to 'seed' new gene diversity into damaged bleached reefs. Finally, it is noteworthy that oil-rich marine photosynthetic algae are the focus of multiple bio-prospecting ventures for new biofuels as alternate energy production [33] and would benefit from banking to reduce genetic drift of important strains.
As we face ever-warming summers around the world, bleaching has the potential to reduce coral numbers and diversity. Only through selective breeding or assisted evolution [54] will many species be able to tolerate the changing oceans, and flexible and thermally tolerant Symbiodinium may be key to this success. Forming a robust and comprehensive cryobank of the world's Symbiodinium is critical for ensuring the maintenance of global coral diversity, will form a important tool for selective breeding experiments to develop more thermally resilient strains of coral, will create a greater understanding of symbiosis and will provide a 'gold-standard' genomics collection allowing full reads for DNA and RNA sequencing for these important cells.