Abstract
Background and Aims Embryonic sporophytes of the moss Aloina ambigua are inducibly desiccation tolerant (DT). Hardening to DT describes a condition of temporary tolerance to a rapid-drying event conferred by a previous slow-drying event. This paper aimed to determine whether sporophytic embryos of a moss can be hardened to DT, to assess how the rate of desiccation influences the post-rehydration dynamics of recovery, hardening and dehardening, and to determine the minimum rate of drying for embryos and shoots.
Methods Embryos were exposed to a range of drying rates using wetted filter paper in enclosed Petri dishes, monitoring relative humidity (RH) inside the dish and equilibrating tissues with 50 % RH. Rehydrated embryos and shoots were subjected to a rapid-drying event at intervals, allowing assessments of recovery, hardening and dehardening times.
Key Results The minimum rate of slow drying for embryonic survival was ∼3·5 h and for shoots ∼9 h. Hardening to DT was dependent upon the prior rate of drying. When the rate of drying was extended to 22 h, embryonic hardening was strong (>50 % survival) with survival directly proportional to the post-rehydration interval preceding rapid drying. The recovery time (repair/reassembly) was so short as to be undetectable in embryos and shoots desiccated gradually; however, embryos dried in <3·5 h exhibited a lag time in development of ∼4 d, consistent with recovery. Dehardening resulted in embryos incapable of surviving a rapid-drying event.
Conclusions The ability of moss embryos to harden to DT and the influence of prior rate of drying on the dynamics of hardening are shown for the first time. The minimum rate of drying is introduced as a new metric for assessing ecological DT, defined as the minimum duration at sub-turgor during a drying event in which upon rehydration the plant organ of interest survives relatively undamaged from the desiccating event.
INTRODUCTION
Vegetative desiccation tolerance (DT, also ‘desiccation tolerant’) is the ability to revive from an air-dry state in equilibrium with an atmosphere at or below ∼50 % relative humidity (RH; Alpert, 2005; Proctor et al., 2007a). Ecological DT is concerned with the effects of desiccation on survival, growth and fitness (as opposed to cellular DT). Although the terminology may vary, two ecological strategies of DT are considered to exist in nature, constitutive DT (CDT) and inducible DT (IDT; Oliver et al., 1998; Mayaba et al., 2001; Stark et al., 2013). While this dichotomy is useful, a more accurate representation is to consider that ecological tolerance lies along a gradient from strictly constitutive to strictly inducible, with the strategy of most DT plant species falling somewhere along this continuum (Pressel et al., 2006; Stark and Brinda, 2015).
The factors of vegetative DT include the rate at which tissues are dried (RoD), the equilibrating water content (WCeq) to which tissues are subjected (which equates to precise water potentials and RH) and the duration that tissues remain continuously desiccated (Duration Dry). These factors were derived by several researchers over the years and summarized in Green et al. (2011). When RoD is combined with WCeq, such a stress is termed ‘intensity of desiccation’, with the understanding that desiccation intensity obscures the individual effects of the two factors (Stark et al., 2013). Once a plant is rehydrated, the less well known physiological states of the tissues proceed from repair/reassembly (in both CDT and IDT species or tissues) to a hardened state that can withstand rapid drying (RD), and finally through dehardening to a dehardened hydrated state (in IDT species or tissues). Although RoD has been studied with respect to recovery/repair (reviewed in Bewley, 1995), little is known as to how the factors of DT relate to hardening and dehardening phenomena.
Rate of desiccation
The rate at which a bryophyte is desiccated has far-reaching effects on recovery and duration, regardless of ecological strategy. In all cases studied, as the RoD is shortened (as ‘intensity’ in most cases), recovery time and extent of damage increase (e.g. Abel, 1956; Schonbeck and Bewley, 1981a, b; Oliver, 1991; Bewley, 1995; Proctor 2003; Pressel and Duckett, 2010; Stark et al., 2013). This inverse relationship has been explored using various rates of drying, with the RD usually defined as a plant passing from full turgor to a desiccated morphology in <1 h and slow drying (SD) variously defined as from 2 to 10 h or even on the order of many days (Greenwood and Stark, 2014). In quantifying the capacity of DT inherent in a species, genotype or plant structure, a useful measure to derive is the minimum RoD a plant structure can tolerate and emerge relatively undamaged upon rehydration. The same would hold for minimum or maximum WCeq and the maximum Duration Dry factors of DT. Oddly, empirical derivations of these measures, when one factor is held constant while the other is varied, are absent. Since RoD appears to be the more important of the factors RoD and WCeq (Watkins et al., 2007 for ferns, with no corresponding study of bryophytes comparing these two factors) and represents the first factor experienced by a plant during drying, we focused on an empirical derivation of minimum RoD in this paper.
Sporophytic desiccation tolerance
Very little information exists on the DT of bryophyte sporophytes, despite the critical evolutionary position of bryophytes, which comprise the first embryonic land plants. Most workers in the field assume that bryophyte sporophytes must have a degree of DT because desiccating events are probable during development, which may take weeks to years depending on the species (Proctor et al., 2007a). The cultured (dehardened) sporophyte of the moss Aloina ambigua exhibits IDT as an embryo, which transitions to CDT during early seta elongation and beyond. This dual strategy was noted as an evolutionary adaptation to greater desiccation stresses incumbent on the sporophyte as it elongates away from the maternal plant (Stark and Brinda, 2015).
Hardening and dehardening to DT
The short-term acclimation to DT is known as hardening, and is a well known if understudied element of bryophyte ecological DT. Hardening differs from the induction of DT by way of SD; the latter phenomenon simply indicates an IDT condition and prepares the structure to tolerate the drying event (the protection not necessarily persisting for further drying events), whereas hardening prepares the structure to tolerate a future desiccating event carried out at any speed. Hardening is demonstrated for bryophyte gametophytes of 12 species by (1) an SD followed by an RD or a harsher drying episode (Dilks and Proctor, 1976; Werner et al., 1991; Marschall and Beckett, 2005); (2) the application of exogenous abscisic acid (ABA) to cultured plants followed by an RD (Werner et al., 1991; Hellwege et al., 1994; Pence, 1998; Beckett, 1999; Mayaba et al., 2001; Marschall and Beckett, 2005; Pence et al., 2005; Oldenhof et al., 2006; Koster et al., 2010; Pressel and Duckett, 2010; Hájek and Vicherová, 2014); and (3) partial dehydration followed by an RD (Schonbeck and Bewley, 1981b; Beckett, 1999; Hájek and Vicherová, 2014), thus validating ‘the hypothesis of the induced drought hardening, pointed out by Höfler (1945 [50]’, quoted from Bopp and Werner, 1993). Protonema of two Sphagnum species were found incapable of hardening to DT, with an SD episode unable to prevent damage incurred by the subsequent RD (Hájek and Vicherová, 2014). During slow drying, ABA is produced and appears to regulate half of the stress proteins expressed in plants responding to an SD (Werner et al., 1991; Wang et al., 2009; Cruz de Carvalho et al., 2014).
The dissipation of the hardened DT condition is known as dehardening or deacclimation, which implies that the hardened hydrated condition is temporal in IDT species and unstudied in CDT species. Bryophytes are known to deharden over a period of days to months (Schonbeck and Bewley, 1981b; Hellwege et al., 1994; Pence, 1998; Beckett, 1999; Hájek and Vicherová, 2014; Stark et al., 2014). Two bryophyte species do not exhibit dehardening to DT (Beckett, 1999; Stark et al., 2012); a strategy of CDT is thus indicated.
Post-rehydration recovery
Upon rehydration, the inrush of water leads to or reflects tissue damage, as evidenced by membrane leakage during the first few minutes (e.g. Oliver et al., 1993; Deltoro et al., 1998), followed by repair and/or reassembly processes that we term here ‘recovery’. The timeline for recovery that returns the plants to near control levels of function varies based on the severity of the previous desiccation, and even though mostly based on field-collected material that may not have been fully deacclimated, appears to be well supported and in the neighbourhood of 12–24 h. A variety of recovery measures suggest this 12–24 h timeline, including oxygen consumption (as returning to control levels and dependent upon prior rate of drying; Bewley et al., 1978), carbon balance (cf. Tuba et al., 1996 and Reed et al., 2012, who found C balance to recover in ∼60 min in Syntrichia; Schonbeck and Bewley, 1981a; Alpert and Oechel, 1985, 1987; Proctor and Pence, 2002; Proctor et al., 2007b; noting that time to positive carbon balance depends upon the duration of the dry period and the equilibrating RH; Hinshiri and Proctor, 1971; Dilks and Proctor, 1974), resumption of control levels of protein synthesis (Oliver, 1991), cellular integrity (Bewley et al., 1993), return to normal appearance of organelles (Bewley and Pacey, 1978; Phillips et al., 2002; Proctor et al., 2007b), chlorophyll fluorescence (Hellwege et al., 1994; Proctor and Pence, 2002; Schlensog et al., 2004; Proctor et al., 2007b; Pressel et al., 2009), vacuole recovery (Proctor et al., 2007b) and cytoskeleton recovery (Pressel et al., 2006). The treatment of plants prior to testing for recovery and the inherent ecological strategy employed by the species under study should each heavily influence results for estimates of recovery, with deacclimated plants expected to segregate into longer recovery (IDT) and shorter/no recovery (CDT) times.
Questions and hypotheses
The embryos of the moss A. ambigua are known to be inducibly DT, with nearly all suffering death following an RD yet experiencing nearly 100 % recovery when subjected to an SD (Stark and Brinda, 2015). Two questions are posed for this species. First, are embryos capable of hardening to desiccation stress? In order to address this question, it is necessary to empirically derive a minimum rate of drying so that we can reliably assign shorter and longer experimental rates of drying. Second, does the rate of desiccation influence the rehydration dynamics of recovery, hardening and dehardening? Neither of these questions has been experimentally addressed for the moss embryo. We hypothesize that (1) embryos can harden to desiccation stress, (2) the rate of prior desiccation is inversely proportional to the length of the recovery, hardening and dehardening intervals, and (3) the minimum RoD is equivalent for shoots and embryos.
MATERIALS AND METHODS
Species description
The genus Aloina is characterized by infolded succulent leaves bearing upright filaments that cover the costa and part of the lamina. Aloina ambigua is distinguished from other members of the genus by having leaves lacking a hairpoint, leaf bases lacking a differentiated margin of thin-walled hyaline cells, and the costa smooth abaxially (Gallego et al., 1999; Norris and Shevock, 2004; Delgadillo, 2007). The sexual condition is rhizautoicous (Stark and Brinda, 2015).
Culture technique
Our culture technique is described in detail in Stark and Brinda (2015). Briefly, shoots from a herbarium specimen [Brinda 2024: USA, Arizona, Mohave County, Grand Canyon-Parashant National Monument, Parashant Canyon, 36.155415N, 113.332383W (WGS84), elevation 774 m, 29 June 2007, on soil] were placed into culture, decontaminated and cloned through a series of subcultures, and grown to sex expression in a plant growth chamber (Percival model E30B, Boone, IA, USA) under a 12-h photoperiod (20 °C lighted, 8 °C darkened; termed ‘recovery settings’) on locally collected sieved and dry-sterilized, pH-neutral sand medium in 35 mm (inner diameter) plastic Petri dishes. The RH inside the Petri dish was stable, near 100 %, while inside the growth chamber RH varied from ∼70 % lighted to ∼90 % darkened, and the plants received 90–120 µmol m–2 s1 photosynthetically active radiation (PAR). Each week, cultures were hydrated, initially with sterile distilled water (early protonema) and then (after 3–4 weeks) with a 30 % Hoagland’s inorganic nutrient solution (Hoagland and Arnon, 1938). Experimental shoots (with sporophytes) were ∼4–7 months old. Fertilization was effected under normal watering conditions, which thoroughly saturated the medium surface. Embryonic sporophytes were detected using a dissecting microscope at 60×.
Water content and RH
Water content (WC) was given in Stark and Brinda (2015) on a dry-weight basis and not repeated here for the same species. RH experienced by the shoot complexes was measured as in Stark and Brinda (2015). Briefly, iButtons (Maxim, San Jose, CA, USA) positioned inside the Petri dish measured RH at regular intervals until RH equilibrated at ∼50 %. Using groups of three embryo–shoot complexes prepared as given in the section Rate of drying, an iButton was placed inside the Petri dish and the dish placed in an electronic desiccation cabinet set to 50 % RH. All activities were carried out within a walk-in environmental room set at 50 % RH and 20 °C. Leaf curling was visually assessed by noting when the leaves were approximately halfway between a full turgor position and a desiccated position in relation to the shoot.
Rate of drying
The minimum rate of drying (RoD) is considered as the duration at sub-turgor (environmental RH 86–96 %) resulting in embryonic survival (sporophytes) or undamaged leaves (shoots) upon rehydration from a desiccating event. The RoD was assessed on embryos at the early embryo phenophase, when the embryo measured 1·5–2·5 mm in length and which were attached to maternal shoots. Each shoot bearing an embryonic sporophyte (embryo–shoot complex, ‘embryos’ or ‘plants’ here) was gently excavated from the culture using blunt forceps to include a small mass of rhizoids at the base of the shoot (for ease of handling, limiting physical injury to the embryos). Embryos were treated in groups of three and cleaned of debris on a microscope slide in droplets of sterile water, and the rhizoid masses were trimmed to be roughly equivalent in size and carefully placed on partially saturated filter paper within a lidded Petri dish for ∼60 s to blot free water from the embryos. The three plants were positioned with rhizoid masses in contact with one another in order to equalize free water availability and drying rates among the plants (Rice, 2012). This group of three embryos was then transferred (grasping the common rhizoid mass) to another Petri dish containing two-ply filter paper (Whatman No. 1 cut to the inner circumference of the Petri dish and the filter paper pre-equilibrated at 50 % RH for at least 24 h). Filter paper was premoistened with a pipette with 0 (unlidded), 0 (lidded), 10, 17·5, 25, 50, 75, 100 or 200 µL of sterile distilled water. These water volumes equated to drying times of ∼15, 25, 70, 130, 210, 540, 1000, 1320 and 2160 min, respectively. The group of three embryos was moved together and positioned at the centre of the filter paper, and the Petri dish was lidded (except for the 0 -µL unlidded treatment) and placed under dim light (2–4 µmol m–2 s–1) at 20 °C in an electronic desiccation cabinet (Totech, Super Dry, Yokohama, Japan) set at 50 % RH within an environmental control room also set at 50 % RH and 20 °C. Drying time was calculated as the time required for shoots to initiate leaf curling in response to desiccation, and coincided with an environmental RH of ∼86 %. Following leaf curling, the plants were allowed to equilibrate at 50 % RH for an additional 12–24 h to reach constant mass at 50 % RH. At this point the embryos were gently separated while dry and planted individually in a Petri dish containing sieved and sterilized locally collected hydrated sand medium within a jar at the recovery settings above. Sporophytes were allowed to develop by watering daily with sterile distilled water as needed. Controls were transplanted without drying into a Petri dish containing locally collected premoistened sand medium (same recovery settings).
Duration of recovery, hardening and dehardening
Recovery time was defined as the post-rehydration time required for an embryo to survive an RD event, i.e. once dried, the duration of rehydration conferring embryonic survival even after an RD event. Hardening was defined as the ability of a shoot or embryo, once rehydrated from a drying treatment, to survive an RD event. Dehardening describes the loss of this ability with prolonged hydration. Once the minimum RoD was determined, treatments of ∼3·5, ∼22 and ∼36 h, coinciding with 25, 100 and 200 µL pipetted onto two-ply filter paper as described above in the section Rate of drying, were administered. Following drying at these three rates and equilibration at 50 % RH for 12–24 h, the embryos were rehydrated and planted upright (similar to plants used in the minimum RoD treatment above) for designated durations. At intervals of 2, 4, 8, 24, 48, 72, 96, 120 and 144 h, the group of three hydrated and upright embryos was excavated, cleaned and blotted as described above in the section Rate of drying, and then exposed to an RD by transferring the embryos as a group of three to an unlidded Petri dish containing two-ply filter paper with no pipetted water, and placed into the electronic desiccation cabinet at 50 % RH, where the plants desiccated in <20 min and were then allowed to equilibrate overnight to a stable WC. Rehydration was carried out as described above in the section Rate of drying, with viability and shoot damage assessed as given in the section Response variables.
Response variables
The viability of the embryos was assessed based on the ability of the embryo to advance through seta elongation to the phenophase capsule expansion (also known as premeiosis). If the sporophyte did not reach premeiosis it was considered as abortive (most abortions occurred prior to seta elongation, but a few occurred during seta elongation). Growth rate of the sporophyte was assessed as the number of days between the phenophases of embryo to seta elongation, and from seta elongation to capsule expansion. Leaf damage was assessed on day 10 (post-rehydration) by observing the shoots at 60× magnification and classifying each leaf as entirely chlorophyllose (value = 1), partially chlorotic (value = 0·5) or entirely chlorotic (value = 0). In order to be classed as entirely chlorophyllose, both the lamina and the lamellar filaments had to be undamaged. The percentage of chlorophyllose leaves was calculated from these values. If a lower leaf was obscured by upper leaves it was not scored. Such visual estimates of leaf damage correlate well to chlorophyll fluorescence parameters (Greenwood and Stark, 2014).
Supplementary genotype
One additional genotype of A. ambigua was tested to confirm the patterns presented in this study by Brinda 2024. This specimen (Brinda 726, USA, Nevada, Clark County, Valley of Fire State Park, Mouse’s Tank Trail, elevation 622 m, 4 March 2006), was cloned to a single genotype and purified in culture as described above for Brinda 2024 and cultured to maturity. The shoot–embryo complexes were then subjected to (1) an RD to confirm the IDT strategy, and (2) an SD followed by a 24-h rehydration period and then followed by an RD to confirm hardening and assess sporophytic growth rate as days to seta elongation (n = 12 for each treatment).
Statistics
Due to the inherent censoring in the leaf damage data, tobit regression models (Tobin, 1958) were used to test for treatment effects. Generalized linear models (GLMs) with binomial errors (logistic regression; Kutner et al., 2004) were used to determine whether rate of drying and/or recovery time affected the likelihood of embryo/sporophyte survival. The sporophyte development times were analysed using Cox proportional hazards regression models (Therneau and Grambsch, 2000). All analyses were performed using R version 3.1.3 (http://www.r-project.org/).
RESULTS
Water content
Water content dynamics were as shown in Stark and Brinda (2015, Figs 1 and 2). In summary, WC declined rapidly during the first 40 min in the RD treatment, reaching very close to equilibration in 2 h; in an SD treatment using 150 µL WC declined much more slowly, reaching equilibrium with 50 % RH after ∼35 h. On a dry weight basis, WC was 12–13 %. For the three hardening treatments (25, 100, 200 µL), RH experienced by the moss is shown in Fig. 1, with the curve inflexion points coinciding approximately with RH falling below 86 % and shoots exhibiting the desiccation morphology of leaf curling in Aloina.
Relative humidity (RH) experienced by shoot–embryo complexes of Aloina ambigua during three experimental drying trials of ∼3·5, 22 and 36-h periods from full turgor to leaf curling (mean ± s.e.; s.e. values not evident are within the symbols) as measured by iButtons placed inside Petri dishes containing two-ply filter paper and sterile water aliquots of 25 (n = 6, 10-min intervals), 100 (n = 5, 30-min intervals) and 200 (n = 6, 60-min intervals) µL, respectively. Leaf curling (where leaves were ∼1/2 appressed to shoot) occurred at arrows (∼86 % RH). Dotted line at 86 % RH coincides with leaf curling and is where metabolism normally ceases (see text).
Embryonic survival as a function of rate of drying (time from full turgor to leaf curling) in cultures of Aloina ambigua. The µL values refer to the volume of water pipetted onto filter paper in the Petri dish to produce the specific drying time, with the first 0 µL unlidded and the second 0 µL lidded; note the unevenness of the x-axis intervals. Values in parentheses are n. Control, undried plants.
Embryonic survival
Embryonic survival was directly related to the rate of drying, with longer drying times resulting in significantly higher survival (Fig. 2, z = 5·873, p < 0·001). In rapidly dried embryos (15 min), survival was only 6 %. However, if embryos were very slowly dried (36 h), survival was 100 % and at the level of control undried embryos (Fig. 2). If minimum RoD is defined as that rate of drying resulting in 80 % embryo survival, this was achieved by allowing the embryos to dry over a period of ≥210 min (3·5 h). This rate of drying is the duration over which the plant is exposed to RH ≥∼86 %, which is the RH at which leaf curling in A. ambigua (and other bryophytes; see Discussion) occurs.
Sporophyte growth rate
Growth rate (as time to seta elongation) was related to the rate of drying, with slower rates coinciding with more rapid drying rates, and faster growth rates coinciding with slower drying rates (Fig. 3, z = 11·47, p < 0·001). This delay in development continued to be evident through the total time to capsule expansion (z = 11·38, p < 0·001), although the average development time from seta elongation to capsule expansion was not different among treatments (Fig. 4), with all treatments taking ∼14 d. Controls and embryos dried at the slowest rate (36 h) took 6 d to reach seta elongation. If minimum RoD is defined as the growth rate within a single day of the controls, this was achieved by allowing the embryos to dry over a period of ≥3·5 h.
Embryonic growth rate (time to reach the seta elongation phenophase) as a function of rate of drying (time from full turgor to leaf curling) in cultures of Aloina ambigua. The µL values refer to the volume of water pipetted onto filter paper in the Petri dish to produce the specific drying time, with the first 0 µL unlidded and the second 0 µL lidded; note the unevenness of the x-axis intervals. Values in parentheses are n. Control, undried plants. Dotted lines at days 6 and 7 are indicated for clarity.
Sporophytic growth rate (time from the seta elongation phenophase to the capsule expansion phenophase) as a function of rate of drying (time from full turgor to leaf curling) in cultures of Aloina ambigua. The µL values refer to the volume of water pipetted onto filter paper in the Petri dish to produce the specific drying time, with the first 0 µL unlidded and the second 0 µL lidded; note the unevenness of the x-axis intervals. Values in parentheses are n. Control, undried plants; dotted line indicates mean of the control.
Embryonic rehydration, recovery, hardening and dehardening
Embryonic survival was affected by both the rate of drying and recovery time but the relationship between these factors was complex (Fig. 5). In the full model, the main effect of drying rate was significant (z = 7·468, p < 0·001) as well as the interaction between drying rate and recovery time (z = 3·481, p < 0·001). However, in this model the main effect of recovery time was not significant (z = –1·732, p = 0·083). The biological interpretation of this is rather straightforward and illustrates that the dehardening process is simply the converse of the hardening process. At recovery times approaching zero (i.e. 1–2 h post-rehydration) the effect of rate of drying on survival was strong since the slowly dried embryos were still maximally hardened. Similarly, at rapid drying rates (i.e. the dotted line in Fig. 5) very little variation in survival is explained by recovery time since the embryos were minimally hardened. The significance of the interaction between drying rate and recovery time is shown by the markedly different slopes of the curves in Fig. 5. This interaction clearly illustrates that for species with a plastic response, such as A. ambigua, precise knowledge of the timing of previous stresses is necessary in order to interpret current performance.
Hardening dynamics of embryos of Aloina ambigua. Survival is shown as fitted polynomial curves. Embryos were dried at three rates (3·5, 22, 36 h, equilibrating at 50 % RH), rehydrated and exposed to rapid drying (RD, 15 min to leaf curling, equilibration at 50 % RH) at 1-, 2-, 4-, 8-, 24-, 48-, 72-, 96-, 120- and 144-h intervals post-rehydration, rehydrated and assessed for survival. Values above the dotted line indicate hardening (6 % = survival of RD dehardened embryos).
For the shoots and embryos to fully rehydrate following a drying treatment took from 2 to 4 h, with the majority of rehydration occurring during the first 60 min. Shoots that were directly immersed in water hydrated in 40–60 min, but this method was not employed because of the wide variation seen in the rate of rehydration with immersed shoots. The centrally located embryo was the last structure, along with the tips of the leaves, to fully rehydrate (and the last structure to desiccate). By subjecting the embryos to an RD at intervals post-rehydration, it is possible to determine the timeline for recovery, hardening and dehardening. A recovery (repair/reassembly) time was not detected for embryos dried for 22 h and longer, even when these embryos were subjected to RD 1 h post-rehydration (i.e. during the rehydration process); no mortality occurred and resumption of embryonic growth was normal. Degree of embryonic hardening was related to the rate of prior drying (3·5, 22 and 36 h), with longer drying times producing significantly longer and more stable hardening times for embryos (Fig. 5).
Shoot survival and leaf damage
Of the 330 shoot–embryo complexes used in the experiment, only three shoots failed to survive, with survival assessed as the capacity to regenerate protonemata over a 30-d period from planting. All three of these non-surviving shoots received an RD treatment. Thus, among the shoots treated with an RD (15 min from full turgor to leaf curling), 30 of 33 survived to regenerate despite incurring heavy damage. Shoot damage was directly related to the rate of drying, with longer drying times resulting in significantly less leaf damage (z = 17·68, p < 0·001; Fig. 6). In rapidly dried shoots, leaf damage was heaviest (14 % chlorophyllose leaves). However, if shoots were very slowly dried (≥22 h), shoot damage was <10 % (>90 % chlorophyllous leaves). If minimum RoD is defined as the rate of drying resulting in 80 % chlorophyllose leaves, this was achieved by allowing the shoots to dry over a period of ∼9 h.
Shoot damage (as percentage of chlorophyllous leaves) as a function of rate of drying (time from full turgor to leaf curling) in cultures of Aloina ambigua. The µL values refer to the volume of water pipetted onto filter paper in the Petri dish to produce the specific drying time, with the first 0 µL unlidded and the second 0 µL lidded; note the unevenness of the x-axis intervals. Values in parentheses are n. Control, undried plants. Dotted lines at 50, 80 and 100 % are indicated for clarity.
Shoot recovery, hardening and dehardening
Leaf damage in response to hardening/dehardening was very similar to that described above for embryos. In the full model, the main effect of drying rate was significant (z = 15·194, p < 0·001) as was also the interaction between drying rate and recovery time (z = –4·916, p < 0·001; Fig. 7). Again, the main effect of recovery time on leaf damage was not significant (z = 0·770, p = 0·441), but this should be interpreted in the same way as above for embryo survival. By assessing leaf damage across three hardening treatments (3·5, 22, 36 h) and subjecting shoots to an RD at intervals post-rehydration, it is possible to detect vegetative hardening: if the degree of leaf damage is less than that seen in a simple RD of cultured control plants followed by rehydration (86 % leaf damage), then hardening to DT is deduced. A recovery (repair/reassembly) time was not detected for shoots; even when shoots were dried during rehydration (1–2 h post-rehydration), some hardening was already evident (Fig. 7). The degree of vegetative hardening was related to the rate of prior drying (3·5, 22 and 36 h), with longer drying times producing significantly longer and more stable hardening times for shoots. However, in all cases shoots did not deacclimate to the level of the undried control plants, indicating that even after a 6-d dehardening period, unlike embryos, shoots retained some hardening to DT.
Hardening dynamics of maternal shoots of Aloina ambigua: leaf damage is shown as fitted polynomial curves. Shoots were dried at three rates (3·5, 22, 36 h, equilibrating at 50 % RH), rehydrated and exposed to rapid drying (RD, 15 min to leaf curling, equilibration at 50 % RH) at 1 -, 2 -, 4 -, 8 -, 24 -, 48 -, 72 -, 96 -, 120 - and 144-h intervals post-rehydration, rehydrated and assessed for leaf damage as the percentage of damaged leaves 10 d post-rehydration. Values above the dotted line indicate hardening (86 % is the percentage of leaf damage for RD-dehardened shoots). Note the descending scale on the y-axis, with 0 % = no leaf damage and 100 % = all leaves damaged. Mean ± s.e., n = 21 for each point.
Supplementary genotype
Embryos of one additional genotype of A. ambigua (from Nevada) exhibited all three of the major patterns found in this study: (1) the inability to withstand an RD when dehardened; (2) the ability to harden to an RD when the RD was administered after a 24-h rehydration period following an SD; and (3) shoot damage that was much more severe after an RD in dehardened plants than an RD that followed an SD (Table 1). The supplementary genotype differed in having a slightly slower embryonic growth rate, exhibiting more damaged leaves following dehardening and hardening treatments, and exhibiting a lower survival of embryos subjected to an RD.
Survival of embryos, growth of embryos and shoot damage exhibited by a supplementary genotype of Aloina ambigua from Nevada, USA. RD, rapidly dried; SD, slowly dried. ‘Dehardened RD’ refers to shoot–embryo complexes that were rapidly dried directly from culture, and ‘Hardened RD’ refers to shoot–embryo complexes that were removed from culture, subjected to slow drying (36 h), rehydrated for 24 h and then subjected to rapid drying. Mean ± s.e. n = 12 for all treatments
| Treatment | Embryo survival (%) | Time to seta elongation (days) | Shoot damage (% chlorophyllous leaves) |
|---|---|---|---|
| Dehardened RD | 0 | – | 5·75 ± 2·91 |
| Hardened RD | 92 | 8·73 ± 0·38 | 40·25 ± 2·78 |
| Undried shoots | – | – | 71·67 ± 1·76 |
| Treatment | Embryo survival (%) | Time to seta elongation (days) | Shoot damage (% chlorophyllous leaves) |
|---|---|---|---|
| Dehardened RD | 0 | – | 5·75 ± 2·91 |
| Hardened RD | 92 | 8·73 ± 0·38 | 40·25 ± 2·78 |
| Undried shoots | – | – | 71·67 ± 1·76 |
Survival of embryos, growth of embryos and shoot damage exhibited by a supplementary genotype of Aloina ambigua from Nevada, USA. RD, rapidly dried; SD, slowly dried. ‘Dehardened RD’ refers to shoot–embryo complexes that were rapidly dried directly from culture, and ‘Hardened RD’ refers to shoot–embryo complexes that were removed from culture, subjected to slow drying (36 h), rehydrated for 24 h and then subjected to rapid drying. Mean ± s.e. n = 12 for all treatments
| Treatment | Embryo survival (%) | Time to seta elongation (days) | Shoot damage (% chlorophyllous leaves) |
|---|---|---|---|
| Dehardened RD | 0 | – | 5·75 ± 2·91 |
| Hardened RD | 92 | 8·73 ± 0·38 | 40·25 ± 2·78 |
| Undried shoots | – | – | 71·67 ± 1·76 |
| Treatment | Embryo survival (%) | Time to seta elongation (days) | Shoot damage (% chlorophyllous leaves) |
|---|---|---|---|
| Dehardened RD | 0 | – | 5·75 ± 2·91 |
| Hardened RD | 92 | 8·73 ± 0·38 | 40·25 ± 2·78 |
| Undried shoots | – | – | 71·67 ± 1·76 |
DISCUSSION
We hypothesized that embryos of the moss A. ambigua should be capable of hardening to a desiccation stress, that the rate of prior drying should be inversely related to the length of the recovery, hardening and dehardening timelines, and that embryos and their maternal shoots should exhibit equivalent minimum rates of drying. In order to properly assign hardening treatments, we needed to empirically derive the minimum rate of drying whereby >80 % of embryos survive and shoots emerge >80 % undamaged, and we hypothesized that embryos and their attached shoots should exhibit similar minimum rates of drying. ‘Hardening’ to DT is defined as the environmentally induced ability to tolerate an RD event in a species or structure that normally is incapable of surviving an RD event, such as the IDT embryonic sporophyte of A. ambigua. We show for the first time that moss embryos are capable of hardening to DT, and that the severity of the prior drying event is directly related to the degree of hardening for both embryos and the attached maternal shoots. If a maternal plant bearing an embryonic sporophyte is dried gradually over a period of 1–2 d, upon rehydration the embryo is immediately able to withstand an RD event that under normal culture conditions would kill the embryo. Further, it maintains a degree of this hardened condition for a period of up to 5 d if the previous drying event was 36 h in length (the most gradual drying period tested). While our first two hypotheses were sustained, our third hypothesis was not – minimum rates of drying were not equivalent between embryos and maternal shoots. While a ≥80 % survival rate for embryos was achievable by drying embryos in 3·5 h, a ≥80 % incidence of healthy leaves was achievable by drying shoots over a 9- to 15-h period, i.e. 3- to 5-fold longer. However, we concede that survival (embryos) and chlorosis (leaves) are not comparable response measures.
Hardening research
In several ways, our findings on embryonic hardening can be extrapolated from the studies of Werner et al. (1991) on the induction of DT in protonemata of the moss Funaria hygrometrica. These authors, after showing that protonema exhibit IDT (inducible DT, unable to withstand an RD but tolerating an SD), demonstrated that (1) both an SD and ABA application hardened the protonema to a subsequent RD, and (2) ABA produced during the SD event stimulated protein synthesis (confirmed in Fontinalis and Physcomitrella; Wang et al., 2009; Cruz de Carvalho et al., 2014), thus implicating the probable mechanism not only for DT but also for hardening to DT. The importance of the rate of drying in inducing DT was thus shown to be both critical to the process of hardening and held in common between IDT bryophytes and IDT vascular plants (Pressel and Duckett, 2010; Cruz de Carvalho et al., 2014). The findings of Werner et al. (1991) were validated in other bryophyte species, including liverworts (Hellwege et al., 1994, 1996; Pence et al., 2005), aquatic mosses (Cruz de Carvalho et al., 2014), desert mosses (Stark et al., 2013, 2014) and Physcomitrella (Greenwood and Stark, 2014), and in partial dehydration experiments (Schonbeck and Bewley, 1981b; Beckett, 1999; Hájek and Vicherová, 2014). However, in CDT species, rate of drying apparently has little or no effect (Zhang et al., 2011; Stark et al., 2012) on the ability to tolerate desiccation, and positive C balance can be attained within 1 h following rehydration (Tuba et al., 1996; Reed et al., 2012). In the future it will be important to discriminate, in plants that exhibit a constitutive photosynthetic response to drying, between an inherently CDT species and an IDT species that has been hardened to DT in the field (Stark et al., 2014), and also to characterize the differences, if any, between the hardened IDT state and the continuous CDT state (in an IDT and CDT species, respectively).
Rehydration, recovery, hardening and dehardening
Data presented here suggest that rehydration-induced repair processes are of non-critical importance to the survival of embryos and shoots following a suitably gradual SD event, thus confirming the experimental evidence of Werner et al. (1991), Pressel et al. (2006, 2009) and Proctor et al. (2007b). Indeed, even when the rehydration process in A. ambigua is interrupted in its later stages with an RD event, no effect is observable on the survival (embryos) and damage (shoots) upon rehydration. We note that embryonic growth rates were delayed by a day under this scenario, which suggests some repair/reassembly may be occurring. It is not possible to compare our results with other studies employing partially dehardened field-collected plants, because, as pointed out by Werner et al. (1991), and later validated through study of dehardening dynamics in Crossidium (Stark et al., 2014), if plants are collected from the field and not thoroughly deacclimated (dehardened) the results are expected to be skewed towards the appearance of a largely constitutive response in a species that may be ecologically inducibly DT (such as Syntrichia ruralis; Schonbeck and Bewley, 1981a).
On the other hand, when plants are subjected to a drying event that is more rapid or of long duration, repair processes may be required in order to attain positive C balance, restore/repair the photosynthetic apparatus (Hellwege et al., 1994; Li et al., 2014), re-establish the cell cycle, translocate metabolites to meristematic regions, reconstitute microtubules and effect DNA repair (Proctor and Pence, 2002). Recent evidence comparing desiccated Bryum argenteum tissues that were dried for 1 week versus 1 year implicate the repair processes of re-establishing protein conformation and resynthesis of thylakoids (Li et al., 2014). In addition, protein transcription patterns associated with rehydration following an SD in Fontinalis were interpreted as implicating repair as a dominant process (Cruz de Carvalho et al., 2014). As noted by Bopp and Werner (1993), ‘the enhanced synthesis of proteins during rehydration is independent of the synthesis of protective proteins during desiccation’. When embryos of A. ambigua were dried at rates of <3 h, the embryos suffered mortality rates of >20 %, and, critically, resumption of embryonic growth was delayed from 1 to 7 d in surviving embryos compared with the SD embryos. These data are consistent with interpretations of lag times in seed germination that implicate DNA repair processes (Osborne et al., 2002; Pammenter et al., 2002), and support a repair-based mechanism to counter the damaging effects of rapid or extended drying through the production of rehydrins that are translated post-rehydration and which play a role in recovery (Velten and Oliver, 2001).
The length and degree of hardening experienced by bryophytes are seldom reported, although for Funaria protonema Werner et al. (1991) estimated the minimum RoD to elicit hardening to be in the neighbourhood of 24 h based upon an ABA incubation time series. In shoot segments of Atrichum, hardening due to a partial desiccation was still in (partial) effect, as measured by ion leakage, 1 week after the hardening treatment (Beckett, 1999). Species of Sphagnum exhibited a degree of hardening 3 weeks following a hardening treatment, the longest period shown (Hájek and Vicherová, 2014). In A. ambigua, SD embryos remain partially hardened over a deacclimation period of 5 d, becoming fully dehardened at 5–6 d following the hardening treatment. After 6 d post-rehydration, some embryos entered seta elongation. How this timeline is affected by consecutive SD events has not been studied.
Rate of drying
Embryos and shoots of A. ambigua differed in their minimum RoD required to result in >80 % survival (embryos, 3 h) or <20 % damage (shoots, 9 h). In the desert moss Pterygoneurum lamellatum, an SD of 24 h was required for dehardend shoots to exhibit <20 % damage (Stark et al., 2013), indicative of significant differences even between two related aridland IDT species. Although ample evidence points to the influence of RoD on recovery time (Bewley, 1995), there is no comparable evidence for the sporophyte generation, and perhaps no reason to believe that the two generations have similar requirements in light of the divergent measures assessed here of survivorship (embryos) versus leaf damage (shoots). The ‘ability of the plant to survive desiccation correlates with the accumulation of carbohydrates’ (Dinakar et al., 2012); the relatively short minimum RoD for embryos indicates that embryos can rapidly transport or convert metabolites to sucrose. Interestingly, the 3·5-h requirement for drying in A. ambigua embryos is similar to the lag time for ABA production during the slow partial desiccation accompanying the transition from aquatic to landform expressions of the liverwort Riccia fluitans (Hellwege et al., 1992). While the shoots of A. ambigua exhibit a clear pattern of IDT, nearly all shoots survived an RD event, indicating that (1) pockets of constitutively protected cells exist in the shoot, which are capable of regenerating protonema, and (2) a fundamental difference exists in the DT strategy of shoots and embryos. This finding is consistent with the view that ecological DT can be viewed along an inducibility gradient, with potential life phase- and age-related differences in the DT response (Pressel and Duckett, 2010; Stark and Brinda, 2015).
Experimental rates of drying vary widely, and aside from an RD normally construed to be <1 h in duration from full turgor to leaf curling, there is no consensus as to what constitutes an SD. Most researchers have followed the conventions described in Bewley (1995), of an RD of <1 h, an SD of 3–4 h and a very SD of up to 12 h. Ideally, experimental drying times will reflect those occurring in nature. Boulder populations of Grimmia in southern California were found to desiccate in 10–17 h (Alpert, 1990). Patch desiccation in Germany (Grimmia) and Arizona (several desert species) desiccated in 6–9 h (Alpert, 1979; Zotz et al., 2000), while desert soil patches of Crossidium in Nevada desiccated at rates from a few hours to 4–5 d, depending on the season and incident rainfall (Stark, 2005). Aside from these observations, most reports of natural drying rates are anecdotal.
The time from full turgor to leaf curling represents an estimate of the rate of drying (time at sub-turgor during a drying event) and seems to correlate well with an equilibrating RH of ∼86 %. In S. ruralis, leaf curling occurs at 86 % RH, and the authors found that only the time to leaf curling (not to WC equilibration) predicted damage upon rehydration (Schonbeck and Bewley, 1981a). Similarly, in Anomodon and Homalothecium, measurable respiration and photosynthesis occurred down to water potentials equating to an equilibrating RH of 86 % (Proctor, 2001), and in Physcomitrella patens an equilibrating RH of 86 % coincided with growth cessation and loss of water content (Buda et al., 2013). Such an RH coincides approximately with the cessation of metabolic activity in several mosses (Dilks and Proctor, 1979), and these observations are consistent with the RH at leaf curling as assessed using iButtons coupled with visual observations here (Fig. 1) and even in green algal lichens (Lakatos, 2011). These correlated findings help in the understanding of how partial dehydration can produce hardening to DT (Schonbeck and Bewley, 1981b; Beckett, 1999), and suggest future experiments on consecutive partial desiccation events that may increase the degree of hardening. Given that net photosynthesis is not possible in mosses below an equilibrating RH of 96 % (Rundel and Lange, 1980), it is likely that the inducible metabolic events resulting in DT and hardening to DT in mosses occur in the window of drying when the bryophyte experiences environmental RH that is falling between 96 and 86 %.
A test of the productivity trade-off hypothesis
Linking DT to slower growth is the hallmark of the productivity trade-off (PT) hypothesis, which is plausible because DT likely entails significant metabolic costs (e.g. synthesis of sugars, proteins, antioxidants) that detract from growth (Oliver et al., 2000; Alpert, 2006). Despite its appeal, experimental tests of the PT hypothesis do not exist. One way to approach a test of the PT hypothesis is to compare the growth rates between a hardened (to DT) structure and an unhardened structure (of a known IDT species or structure) in the same genotype or species. The prediction is that the growth rate of the hardened structure should be slower than that of the unhardened control, because the hardened structure has experienced the induction of DT with associated metabolic costs, and a delay in growth resumption is expected to repair damage incurred or recover/set into action events related to the desiccating event. In the present experiment, we can compare the hardened versus unhardened embryos of A. ambigua, assessing growth rate as the time for the embryo to reach the phenophase of seta elongation. From Fig. 3, it is apparent that these two treatments reach seta elongation at the same time (6 d). These data, while scant, control for genetic variation and age, do not support the PT hypothesis, and, at a minimum, indicate that the time duration from embryo to seta elongation does not trade off with DT. During the normal course of sporophyte maturation, A. ambigua sporophytes pass from IDT embryos to CDT in the seta elongation phase. Since the CDT state is generally construed as metabolically more expensive to maintain, the rapid expansion of the seta in this and other species is also puzzling.
Modelling DT in the bryophyte clade
An elegant model/schematic diagram of DT is given in Green et al. (2011), which illustrates the ‘environmental and organism drivers of DT in lichens and bryophytes’. Factors of drying per se consist of RoD, water content (RHeq, as ‘depth’) and Duration Dry (as ‘length’). For sporophytes, RoD may be of senior importance among these factors; at least in A. ambigua, the RoD is linked to sporophyte survival, growth rate, fitness (Stark and Brinda, 2015) and hardening capability. In this model of Green et al. (2011), the possibility of hardening to DT is indicated with a previous slight drying event, and the length of the recovery interval is given as the first phase of rehydration. If the hydrated states of hardening and dehardening are added to the model, then we have three factors of drying preceding three hydrated physiological states, in the sequence of recovery → hardening → dehardening. Organism effects include the inherent degree of DT of the study species, which will fall under the two main ecological strategies, CDT and IDT, with the understanding that these two strategies are not absolute but grade into one another. How the factors of DT, separately and in combination, influence the physiological states of hydration and the fitness of the organism are important questions in plant DT.
Based on the experimental use of acclimated plants, the seemingly rapid response, upon rehydration, for C balance and photosynthesis led scientists to believe that the bryophyte clade was constitutive, indeed defining a basic distinction between DT bryophytes and DT vascular plants (Oliver and Bewley, 1997; Oliver et al., 1998; Toldi et al., 2009; Farrant and Moore, 2011; Dinakar et al., 2012). This distinction is proving to be untenable, as at least as many IDT bryophyte species as CDT species exist (Stark et al., 2013; Hájek and Vicherová, 2014), and as more studies utilize fully deacclimated plants it will probably become clear that ecological strategy is not a strict dichotomy but rather exists along a gradient of ecological inducibility (Pressel and Duckett, 2010; Stark and Brinda, 2015). All three desert bryophytes tested to date are IDT, in opposition to the prevailing view that aridland bryophytes should be under selection pressure towards CDT (Marschall and Beckett, 2005). Finding a common strategy among desert species, Physcomitrella, Funaria, protonemata of various bryophytes and the aquatic Fontinalis lends credence to the hypothesis, voiced by Werner et al. (1991) and subsequently validated in divergent species (Wang et al., 2009; Pressel and Duckett, 2010; Cruz de Carvalho et al., 2014), that DT vascular plants and DT bryophytes probably do not differ as much as previously thought in their mechanisms of tolerating desiccation. A key element will be the comparison of ecological strategy and cellular mechanisms between inherently IDT and CDT species, an experiment well suited to be carried out in bryophytes.
ACKNOWLEDGEMENTS
We thank the US National Park Service (Great Basin Cooperative Ecosystem Studies Unit, Cooperative Agreement Number H8R07060001) for allowing collections within the Grand Canyon-Parashant National Monument. This work was supported in part by the US National Park Service, Great Basin Cooperative Ecosystem Studies Unit (Cooperative Agreement No. H8R07060001), a Faculty Sabbatical Leave from UNLV for L.R.S. and graduate assistantship support through UNLV for J.L.G. and T.A.C.
LITERATURE CITED
