Moss survival through in situ cryptobiosis after six centuries of glacier burial

Cryptobiosis is a reversible ametabolic state of life characterized by the ceasing of all metabolic processes, allowing survival of periods of intense adverse conditions. Here we show that 1) entire moss individuals, dated by 14C, survived through cryptobiosis during six centuries of cold-based glacier burial in Antarctica, 2) after re-exposure due to glacier retreat, instead of dying (due to high rates of respiration supporting repair processes), at least some of these mosses were able to return to a metabolically active state and remain alive. Moss survival was assessed through growth experiments and, for the first time, through vitality measurements. Future investigations on the genetic pathways involved in cryptobiosis and the subsequent recovery mechanisms will provide key information on their applicability to other systematic groups, with implications for fields as divergent as medicine, biodiversity conservation, agriculture and space exploration.

after thawing in the laboratory, showed regeneration of moss shoots 15 . These studies were interpreted as cases of cryptobiosis, but did not involve direct measurements of moss metabolic vitality.
Mosses are major components of plant communities in terms of diversity and biomass in high latitude ecosystems in the High Arctic and in Antarctica, where macroscopic vegetation is dominated by cryptogams 16,17 . Mosses are poikilohydric organisms with high degree of phenotypic plasticity 18 and have well-developed features allowing them to tolerate stress 15 and thrive in environments subject to change and disturbance 16 . Among the key traits allowing mosses to persist in harsh environments, one of the most important is tolerance to desiccation and drought 18 , also contributing to their ability to resist freezing (as during the Antarctic winter) 19 .
In this study, measurements of metabolic activity through chlorophyll a fluorescence (as indicators of vitality and survival) and growth chamber experiments applied to mosses exposed by recent glacier retreat at Rothera Point (maritime Antarctica) present evidence of: a) in situ moss survival for over more than 600 y in the natural habitat through cryptobiosis, and their subsequent spontaneous recovery to active metabolism after glacier retreat, b) persistence of active metabolism and growth in surviving exhumed mosses, c) regeneration with development of green stem apices from moribund stems of exhumed mosses.
Gaertn. and Sanionia uncinata (Hedw.) Loeske, were collected from six sites in the glacier front (Fig. 1A,B) and foreland: five samples of exhumed mosses (n. [1][2][3][4][5], plus an extant moss (n. 6, located at 16.8 m from the ice margin in 2011) used for comparison ( Table 1). The sample n. 1 was emergent during the week of the collection, and the sample n. 2 was located less than 1 m from the glacier front and likely exhumed less than one year previously ( Table 1). The remaining three samples (n. 3-5, Table 1) were located close to the glacier front (between <1 m and 10.1 m from the ice margin in 2011) and exhumed between six and 18 years ago (Table 1). All exhumed samples were from in situ habitats in the field, with exceptional preservation and no damage apparent to the gametophytes which exhibited intact rhizoids, stems and leaves. No sporophyte tissue was present. All samples lacked any sign of regrowth, with green lateral branches and stem apices being absent. The calibrated age (see methods) of three of the most recently exhumed moss (n. 1, 3, 4) was determined by radiocarbon dating and ranged between 640-526 and 626-510 cal years BP (Table 1).
Based on field observations and examination in the laboratory with a dissecting microscope (after sample collection, and prior to the growth experiment), two exhumed mosses (n. 1, 2) and the extant sample (n. 6) showed persistence of visibly greenish material with green leaves ( Fig. 2A,B) lacking any decomposition processes and/ or development of moulds. The other three exhumed samples (n. [3][4][5] were characterized by blackish leaves, and/ or their desiccation (Table 1). From these observations, we suggest the two recently exhumed and extant mosses were biologically active populations whereas the other three exhumed mosses were moribund (Table 1).

Vitality
Measurements. Moss vitality was tested by analyzing the metabolic activity of the gametophores using Chlorophyll a fluorescence with reference to a) the values of Fv/Fm (variable fluorescence/maximum fluorescence maximum) (Table 1), and b) the shape of the induction curve (Fig. 3). We considered the gametophytes to be metabolically active when Fv/Fm >0.2 20,21 . Applying this threshold, metabolic activity (compatible with life/vitality) was detected for all the three mosses suggested to be biologically active based on visual inspection (n. 1, 2, 6). The three mosses suggested to be moribund (n. 3-5) exhibited values of Fv/Fm slightly higher than the threshold of 0.2, also indicating metabolic activity, but at a poor level.
Analyzing the fluorescence induction curve shape, the progressive flattening observed comparing the extant moss specimen with the two recently exposed moss clumps indicated a progressive decrease of activity. Conversely, the flattened shape of the fluorescence induction curve of the three mosses suggested them to be moribund (samples n. [3][4][5] and indicated a lack of activity. Growth Experiment. During the growth experiment, the three metabolically active samples (based on their Fv/Fm values; n. 1, 2, 6) remained alive (sample n. 2, Fig. 2A,B), with the extant moss (n. 6) and the recently  Table 1) with the position of the glacier boundary in 1993 (blue line), 2005 (yellow line) and 2011 (red line) (map created by CorelDraw 6.0). exposed sample of S. uncinata (n. 1) exhibiting growth ( Table 1, Fig. 2F,H), and growth occurred from the stem apex in both the latter cases (without any protonemal or rhizoid initial growth). Among the moribund mosses, sample n. 3 became marcescent and covered by fungal hyphae, and did not show any regeneration (Table 1), whereas the other two moribund samples (n. 4, 5) exhibited regeneration ( Fig. 2C-E,G), with growth of leaves at the branch apices in existing gametophyte shoots (Table 1).

Discussion
Cryptobiosis has already been demonstrated for mosses in polar environments, based on in vivo observations 14 and in vitro regeneration experiments 14,15 . Our data, based on fluorescence measurements, corroborate these previous studies. Cryptobiosis in our study is indicated by long-term survival for six centuries, but the study's true novelty is the demonstration, based on vitality measurements as well as on growth experiments, of the ability of entire moss individuals to withstand six centuries of cryptobiosis and, at its cessation, to spontaneously recover to active metabolism and survive in situ following glacier retreat. Indeed, the chlorophyll a fluorescence measurements indicated the occurrence of metabolic activity (compatible with life) for all the three mosses hypothesized to be active based on visual inspection (n. 1, 2, 6). Their Fv/Fm values were lower than would typically be recorded from extant field samples (ranging between 0.6 and 0.8) 21 . However, it is well known that repeated freeze-thaw cycles decrease Fv/Fm, especially when mosses are exposed to light 22 , as would be the case for these mosses re-exposed by the glacier retreat during the austral summer (with 24 hours of light per day). These results are confirmed by the fluorescence induction curve (Fig. 3), that shows a decrease of activity from the extant moss and the two recently exposed mosses. This is similar to the report of Rutten and Santarius 23 when testing age-related differences in frost tolerance of mosses subject to experimental freezing, in particular, the data obtained from the extant moss (sample n. 6 All vitality analyses indicated, at best, poor metabolic activity for the mosses with longer re-exposure times and suggested to be moribund on the basis of visual observations (n. [3][4][5]. Using the extant moss specimen for comparison, our data indicated the persistence of active metabolism and capacity for regrowth of the very recently re-exposed mosses (n. 1), whereas survival and active metabolism were observed for the less recently exhumed moss clumps (≤1 y) (sample n. 2). The mosses that had been exposed for longer intervals were considered moribund as indicated by poor or lack of metabolic activity, but showed regeneration in the growth experiment, as noted in some previous studies 14,15 .
Recovery processes leading to resumption of normal metabolic activity after cryptobiosis are often characterized by high rates of respiration repair processes that, in some cases, can be lethal 24 , especially if mosses are exposed to light, due to photoinhibitory damage 22 . The intensity and frequency of these recovery processes may explain the different ability of mosses to survive and grow after different periods of re-exposure. As time progresses (>1 y) after re-exposure, the mosses will face repeated seasonal dormancy and adverse conditions 25 . Therefore, accumulated damage with the recovery processes will increase with time, explaining why longer re-exposure time may be associated with the observed lower vitality.
Future perspectives may concern investigations on the genetic mechanisms involved in cryptobiosis and the subsequent recovery mechanisms, which would also provide information on the potential exportability and applicability of these mechanisms to other systematic groups (including humans), with potential relevance for medicine (e.g. organ banking through the cryopreservation of complex tissues 26 ), biodiversity conservation (germplasm, embryogenic cells of both gymnosperms and angiosperms 27 ), agriculture (cultivars of major food crops 28 ) and space exploration (dormancy during space travel 29 , use of anhydrobiosis and cryobiosis to protect organisms exposed to space and exoplanetary environments 30  , and sporadic mosses. The study site was located at the southern limit of the Wormald Ice Piedmont Glacier, locally known as the "ice ramp", and located between 10 and 110 m a.s.l. This glacier front has been monitored since 1989 and has receded rapidly in recent decades, with ablation rates (surface lowering) of 0.32 m water equivalent between 1989 and 1997, and a frontal retreat that increased from 0.   (Table 1). Three samples (n. 1, 3, 4) collected in February 2011 (Table 1) were also subsampled for radiocarbon dating. 14 C Dating. Age determination of samples followed a standard 14 C protocol, as used also for the reconstruction of the glacier dynamics during the Medieval Warm Period 31 . The moss samples selected for radiocarbon dating (samples n. 1, 3, 4) were kept frozen between collection and being sent to Beta Analytic laboratories (Miami, USA). These, moss samples were subjected to an acid-alkali-acid pre-treatment according to Beta Analytic standard procedures. After pre-treatment, samples for radiocarbon dating were prepared for AMS by converting them into graphite. Calibrated ages were calculated with the software OxCal 4.2 using the SHCal13 14 C Southern Hemisphere atmosphere dataset. Radiocarbon age data are reported as conventional radiocarbon years BP ( 14 C yr BP) ±σ and as calibrated age ranges with a 2σ error (95.4%) (cal. yr BP; relative to AD 1950).
Chlorophyll a fluorescence test to assess moss vitality. Chlorophyll a fluorescence provides a measurement of the metabolic activity of the gametophytes 21, 32 , with tissue having an Fv/Fm (variable fluorescence/maximum fluorescence) of 0.2 or higher 20, 21 considered to be active. To assess the vitality of the six moss specimens, chlorophyll a fluorescence measurements were carried out in the laboratory with a portable Plant Efficiency Analyzer (PEA, Hansatech Instrument Ltd., Norfolk, UK) fluorometer. Three sub-samples of each of the six frozen moss samples were placed in Petri dishes and placed at ambient temperature (20 °C). Once thawed, all sub-samples experienced 24 h of dark-adaptation and hydration with deionized water 19,22 . From each sub-sample five leaves were randomly detached and each leaf was placed in a different clip with shooters. Assays were performed after an additional dark period of 15 min 22,33 . Measurements were performed with a light intensity of about 3,000 mol m −2 s −1 provided by an array of six light-emitting diodes (peak wavelength at 650 nm) focused onto the sample to provide homogeneous illumination over the exposed area. Fluorescence values were recorded at different rates depending on the different phases of the induction kinetics, in order to provide the best time resolution. Initially data were sampled at 0.01 ms (millisecond) intervals for 2 ms (interval for initial data acquisition to determine both the minimum fluorescence value (F 0 ) value as well as to monitor the initial rapid rise of the chlorophyll fluorescence signal); then slower acquisition rates of 1 ms and 100 ms were used for 2-1000 and 1000-5000 ms assay intervals, respectively, appropriate to follow the decrease of the kinetics of the fluorescence signal. Overall, for each 5 sec (second) assay 1258 data points were recorded. The F 0 parameter (minimum fluorescence) represents the emission by excited chlorophyll a molecules in the antennae structure of Photosystem II (PSII) at time 0. Using a PEA fluorometer, F 0 was calculated for the initial experimental data recorded at the onset of the illumination in the range of 80-120 μs; these data were analyzed using least squares regression by a linear function and then extrapolated to time 0, whose fluorescence signal is considered the F 0 value 34 Table 1.
The kinetics of the light reaction provides an assessment of the recovery of the electron transport in PSII by analyzing the shape of the fluorescence induction curve 35 . To identify all steps in the induction curve a logarithmic presentation of the time axis was used 34 . According to the standard procedure adopted when using PEA fluorometers, the time axis of the induction curve started from 50 μs 34 . Normally the induction curve is characterized by four steps (named O, J, I, P), however samples subject to freezing exhibited different sensitivity of young vs. mature vs. old leaves, with resultant changes of the induction curve shape 23 . To achieve further confirmation of the vitality of the samples analyzed, and to test for differences linked to their different exposure ages, the induction curves for each assay were recorded and logged.
Growth experiment. The growth experiment utilized the natural substrate from beneath moss populations at Anchorage Island (close to Rothera Point) mixed with high quality commercial litter (Tercomposti s.r.l. Brescia). Twelve glass cups were filled each with 9 g of this soil substrate. The 12 filled cups were sterilized at 120 °C for 15 min, transferred into sterilized Petri dishes, sealed with parafilm and placed in a growth chamber (CV-36L5, Percival Scientifics Series 101) from 17 January 2014 to 17 March 2014. In the growth chamber they were kept under controlled conditions at a fixed temperature of 15 °C, with 16 h of light (17 W cool-white fluorescence lamps, light intensity of 64 µmoles/m 2 /s) and 8 h of darkness 14,15 . After two months in the growth chamber all 12 Petri dishes with soil substratum were completely sterile. Six were then used for the moss samples and 6 as controls. On 18 March 2014 the moss samples were transferred into the Petri dishes with the soil substratum: in each Petri dish each moss sample was divided into three sub-samples. Then the 12 Petri dishes were returned to the growth chamber and maintained under the same controlled conditions 14,15 . Once per week each Petri dish received 5 ml of deionized water (misting each specimen with a handheld spray bottle), which had previously been sterilized for 15 min at 120 °C 14 . All samples were placed in shallow trays and repositioned every 15 d to minimize effects of any variations in light intensity within the chamber 14 . The contents of the Petri dishes were inspected visually (including both the moss samples and the six soil control plots) and regularly observed under a dissecting microscope 15 between 19 May and 2 December 2014.