Tolerance to environmental desiccation in moss sperm


  • Erin E. Shortlidge,

    1. Department of Biology and Center for Life in Extreme Environments, Portland State University, PO Box 751, Portland, OR 97207-0751, USA
    Search for more papers by this author
  • Todd N. Rosenstiel,

    1. Department of Biology and Center for Life in Extreme Environments, Portland State University, PO Box 751, Portland, OR 97207-0751, USA
    Search for more papers by this author
  • Sarah M. Eppley

    1. Department of Biology and Center for Life in Extreme Environments, Portland State University, PO Box 751, Portland, OR 97207-0751, USA
    Search for more papers by this author

Author for correspondence:
Erin E. Shortlidge
Tel: +1 503 863 9979


  • Sexual reproduction in mosses requires that sperm be released freely into the environment before finding and fertilizing a receptive female. After release from the male plant, moss sperm may experience a range of abiotic stresses; however, few data are available examining stress tolerance of moss sperm and whether there is genetic variation for stress tolerance in this important life stage.
  • Here, we investigated the effects of environmental desiccation and recovery on the sperm cells of three moss species (Bryum argenteum, Campylopus introflexus, and Ceratodon purpureus).
  • We found that a fraction of sperm cells were tolerant to environmental desiccation for extended periods (d) and that tolerance did not vary among species. We found that this tolerance occurs irrespective of ambient dehydration conditions, and that the addition of sucrose during dry-down improved cell recovery. Although we observed no interspecific variation, significant variation among individuals within species in sperm cell tolerance to environmental desiccation was observed, suggesting selection could potentially act on this basic reproductive trait.
  • The observation of desiccation-tolerant sperm in multiple moss species has important implications for understanding bryophyte reproduction, suggesting the presence of a significant, uncharacterized complexity in the ecology of moss mating systems.


The majority of eukaryotic male organisms produce gametes in the form of sperm, which are generally motile and must travel before arriving at a receptive female. Within eukaryotic organisms, sperm may experience a diversity of situations before reaching the female egg, including free release into the environment (Franzen, 1956; Rouse & Jamieson, 1987; Lord & Russell, 2002; Vanderpoorten & Goffinet, 2009). Land plants require sperm to move some distance, ex planta, to fertilize the female egg and form a zygote, but they do so via diverse pathways. For the majority of tracheophytes, including gymnosperms and angiosperms, male sperm are contained during transport by pollen grains, which are carried by wind or insects across the terrestrial landscape to compatible female sex structures. Upon germination, the pollen grain grows a protective pollen tube that delivers the sperm internally to the egg (Brewbaker & Kwack, 1963; Taylor & Hepler, 1997; Lord & Russell, 2002), although in some gymnosperms (e.g. cycads) flagellated sperm travel internally to fertilize the egg after pollination. Alternately, the nonseed tracheophytes, including pterophytes (ferns) and bryophytes (mosses, liverworts, hornworts), release sperm onto the landscape for water dispersal (Grout, 1900; Garbary et al., 1993; Vanderpoorten & Goffinet, 2009). The journey of these sperm is unique as the sperm are neither released into a body of water nor protected by a pollen grain, but instead are freed onto the terrestrial landscape in an often ephemeral film of water (Muggoch & Walton, 1942; Paolillo, 1981).

Bryophyta, or mosses, are the most species-rich of the three phyla of bryophytes, and the most widely distributed of all seedless tracheophytes (Shaw & Goffinet, 2000). Male reproductive organs (antheridia) of sexually reproducing mosses dehisce lipid-rich sperm masses that contain numerous biflagellated sperm for dispersal into the environment (Paolillo, 1981). However, little information is known about the fate and journey of the moss sperm, a presumably sensitive life stage, once they are freed from the sperm mass to disperse and arrive at a receptive female. The common conjecture has been that the sperm of mosses are dispersed via mass flow of water or by water droplets from splash cups, and as a result dispersal distances are relatively small (Clayton-Greene et al., 1977; Wyatt, 1977; Longton & Greene, 1979). This limited view of sperm dispersal comes under scrutiny with further recent research indicating that moss sperm may disperse an order of magnitude farther than previously believed (Van der Velde et al., 2001; Bisang et al., 2004). It has been long speculated that microarthropods are involved in sperm dispersal, as they are regular constituents of moss canopies worldwide (Gibson & Miller-Brown, 1927; Muggoch & Walton, 1942; Gerson, 1969). This idea was tested and supported in laboratory experiments conducted by Cronberg et al. (2006), using the moss Bryum argenteum in conjunction with added springtails and mites, which are both common inhabitants of moss canopies (Lindo & Winchester, 2006). These data strongly suggest that a more comprehensive perception of the ecology of moss mating systems is necessary, particularly as over 60% of mosses have separate sexes (Wyatt & Anderson, 1984), rendering potentially significant distances between male and female mosses a distinct possibility.

Similarly, our understanding of the relationship between sexual reproduction in mosses and environmental stress is limited (Convey & Smith, 1993; Stark et al., 2000), despite the fact that mosses are globally prevalent and well known for their ability to tolerate environmental stress (Seel et al., 1992; Lovelock et al., 1995; Meyer & Santarius, 1998; Minami et al., 2003; Oliver et al., 2005; Clarke & Robinson, 2008). The majority of research previously conducted on moss sperm focused on sperm architecture and the composition of the sperm mass (Paolillo, 1977, 1979, 1981; Bernhard & Renzaglia, 1995; Renzaglia et al., 2001), the mechanism of initial sperm release (Paolillo, 1975), and the behavior of the released sperm mass upon encountering the air–water interface (Muggoch & Walton, 1942; Paolillo, 1981). Working forward from these studies in combination with what we know about the daily stresses of many mosses, it seems pertinent to investigate the effects of stress on the reproductive cycle of mosses. To our knowledge, only one previous study has examined the impacts of environmental stress on moss sperm (Rosenstiel & Eppley, 2009). This study found that sperm cells of the geothermal moss Pohlia nutans exhibit remarkable thermotolerance, with cells maintaining integrity at temperatures of 60°C. Whether this property is unique to geothermal populations of P. nutans, or more broadly observed in moss sperm, remains unknown. If moss sperm disperses over long distances and these distances include potentially stressful environments of extreme temperatures, sun-exposed canopies, and microarthropod bodies, selection may increase sperm tolerance to those environments, thus enhancing reproductive success of males with stress-tolerant sperm.

All organisms require water for normal metabolic activity (Hochachka & Somero, 2002), yet taxa from a variety of independent lineages are able to survive extreme dehydration (anhydrobiosis; Giard, 1894). In plants, this ability is widespread at reproductive stages (e.g. pollen, seeds, spores), but less common in vegetative stages (primarily found in algae, mosses and lichen; Alpert, 2000). Surviving the loss of cellular water requires having a suite of stress response traits that are associated with preventing oxidative damage, stabilizing macromolecules, and maintaining membrane integrity (Hoekstra et al., 2001; Hochachka & Somero, 2002). Plants vary according to the degree to which they can undergo these responses, on whether they rely on constitutive or induced protection, and on whether they can repair damage upon recovery (Alpert, 2000). Several characteristics of plant response to extreme dehydration are common among taxa, including cellular recovery from stress being dependent on the rate of water loss (e.g. Schonbek & Bewley, 1981; Oliver et al., 1997), and the ability to accumulate sugars during the latter stages of dehydration (e.g. Bewley et al., 1978; Hoekstra & Vanroekel, 1988; Koster & Leopold, 1988; Sun et al., 1994; Hoekstra et al., 2001).

Of the possible and probable abiotic stresses encountered by moss, water stress could be argued the most physiologically and ecologically relevant to the microworld of mosses as they are poikilohydric plants (Proctor et al., 2007) and thereby are unable to regulate internal water relations separate from that of their surrounding environment. In some habitats, mosses may experience dehydration events that range from seasonal, to monthly, to daily (Alpert, 2000; Oliver et al., 2005). Most mosses are believed to possess some degree of desiccation tolerance (Proctor & Pence, 2002) and, like other organisms that tolerate extreme water stress, can dehydrate to exceedingly low water content (5–10% of their DW) and regain physiological function upon rehydration. Gametophytic tolerance to desiccation in mosses has been well studied (Bewley, 1973, 1979; Dilks & Proctor, 1979; Stark et al., 2005; Proctor et al., 2007), and mosses are quickly emerging as an important model system for understanding the molecular and cellular bases of desiccation tolerance (Wood & Oliver, 2004; Cove et al., 2006; Cuming et al., 2007). Nonetheless, only a few studies have focused on desiccation tolerance in moss reproductive structures, including on sporophytes and asexual propagules (Oliver et al., 2000a; Proctor, 2000; Stark, 2002; Rowntree et al., 2007). A remaining question is whether desiccation tolerance is also present in the sperm cells of mosses. Desiccation tolerance in moss sperm would significantly challenge the assumption that moss mating systems are dictated by available free water.

The objective of this study was to investigate how dehydration–rehydration events impact sperm cell integrity in three moss species. We chose three cosmopolitan species with separate sexes that all exhibit desiccation tolerance in their gametophyte phase (Proctor et al., 2007; Wood, 2007): Bryum argenteum (Bryaceae), Campylopus introflexus (Dicranaceae), and Ceratodon purpureus (Ditrichaceae). Specifically, we examined whether environmental desiccation tolerance in moss sperm is influenced by the rate at which cells dehydrate; the impact of exogenous sugars (sucrose) on environmental desiccation tolerance of moss sperm; and the potential for individual or species-level variation in moss sperm tolerance to environmental desiccation.

Materials and Methods

Study species and selection of sexual structures

Three moss species were used for experiments described in this study. Ceratodon purpureus (Hedw.) Brid. and Bryum argenteum Hedw. are nearly cosmopolitan species, with separate sexes (Richmond & Hunter, 1990; Shaw & Beer, 1999; Smith & Convey, 2002; Turmel et al., 2006). Campylopus introflexus Brid. also has separate sexes and is native to the southern hemisphere, but was relatively recently introduced to North America (Gradstein & Sipman, 1978; Frahm, 1980). All three of these moss species have been found experimentally to be tolerant to desiccation events (Equihua & Usher, 1993; Robinson et al., 2000; Proctor, 2001; Nabe et al., 2007).

Plants were collected in 2009 for this study from various locations. Three C. purpureus populations were collected in the Portland, Oregon, greater metropolitan area, from northeast Portland, the Portland State University campus, and North Plains, Oregon, with populations a minimum of 5.8 km apart. The B. argenteum plants were from southern Arizona, northeast Portland, southwest Portland, and the University of Kentucky campus. C. introflexus plants were collected from Lassen Volcanic National Park (LVNP) (a subset of which were stored dry for transport and longevity, until we were successful at growing C. introflexus to sexual maturity in the glasshouse) and the central Oregon coast. Plants were grown in 6.4 × 6.4 cm pots on a substrate of 2 : 1 propagation grade sand and peat moss. Individual plants were field-collected and then grown either from single spore or from gametophyte cuttings. Mosses grown from spore were maintained in Adaptis 1000 Conviron growth chambers (Pembina, ND, USA) through the early juvenile stage (protonemal stage) and then transferred to a glasshouse at Portland State University (PSU). Mosses were grown in a ‘common garden’ setting to reduce variation as a result of previous environmental conditions of the parental plant material, allowing for the partitioning of genetic vs environmental variation (Shaw, 1986). Male moss plants expressing mature reproductive organs (antheridia) were selected for sperm extractions. Mature antheridia typically develop a yellow-orange color, enabling them to be differentiated from immature, often green, antheridia (Bold, 1987). Senescent antheridia are often dark or wrinkled in appearance and can also be distinguished from mature antheridia (E. E. Shortlidge, pers. obs.). Individual mature male perigonia (clusters of antheridia with surrounding modified leaves) were identified with a Leica DMZ 9.5 stereo microscope and removed from the plants with sterile forceps for sperm extraction.

Sperm extraction and cellular integrity assessment

Sperm extraction from all species began with the selection of three to 20 perigonia per plant as described in Rosenstiel & Eppley (2009). Briefly, all sperm extractions and rehydrations were performed using locally collected rainwater (average (± SEM) pH: 6.0 ± 0.09) that was frozen and filtered and supplemented with tetracycline (20 μg ml−1) to deter bacterial growth (Rosenstiel & Eppley, 2009). Perigonia were placed in a small pool (c. 10 μl per perigonium) of rainwater on multiple glass microscope slides and gently agitated with curve-tipped forceps to encourage the sperm masses to release from the antheridial jackets. Once the sperm masses were dehisced (as evident by visible whitish masses emerging from the antheridia), the slides were placed in a hydration chamber for c. 30 m to encourage continued sperm mass release and subsequent sperm cell release (see Paolillo, 1981 for review). The rainwater and released sperm were then pipetted from the slides into microcentrifuge tubes, briefly centrifuged at 3000 g in order to pellet any fragmented plant material, and the resulting sperm cell suspensions were used for experimentation. For all experiments other than when testing for genetic variance among individuals, each sample contained representative perigonia from multiple individuals to minimize bias or effects of individual inequalities. The number of intact sperm cells was assessed using Cell Vu DRM-600 Sperm Counting Chambers (Millennium Sciences Inc., New York, NY, USA) at ×400, under phase contrast, using a Leica DME compound microscope. Observations of cellular integrity (intact cells) by Cell Vu chambers before and after treatments were initially confirmed by ‘live-cell’ staining (SYTO® 12; Molecular Probes, Invitrogen, Eugene, OR, USA), coupled with an Olympus BX60 epifluorescence microscope. Consistent with results from phase-contrast, epiflourescence confirmed that a fraction of moss sperm cells, post-desiccation, were found to maintain cellular integrity. Further, results of flow cytometry cell counts (Millipore), and nuclear DNA staining of sperm cells with 4,6-diamidino-2-phenylindole (DAPI; Molecular Probes, Eugene, OR, USA), before and after sonic disruption (Cole Parmer, GEX130, 40% power, 30 s), verified cellular integrity under phase-contrast. Phase-contrast was found to be a rapid and robust way to determine sperm cell integrity, without the use of dyes; replicate counts (a minimum of two) were taken for each sample at each time point. In all, over 500 sperm samples were assessed over the duration of the experiments. The use of Cell Vu Sperm Counting Chambers has been found to be reliable in giving reproducible accurate sperm cell counts (Mahmoud et al., 1997; Lu et al., 2007).

As our measure of sperm tolerance to desiccation, we report sperm cell integrity rather than sperm motility, which we have used previously (Rosenstiel & Eppley, 2009), for two reasons: the ability of bryophytes to regain cellular integrity after desiccation has been well studied (Oliver et al., 2005); and, in further assessing moss spermatozoa systems, we have become aware of high variance in sperm motility; we have documented bimodal distributions within individuals for all three species in this study (E. E. Shortlidge, T. N. Rosenstiel & S. M. Eppley, unpublished). We believe that future studies should examine the relationship between sperm motility and viability in bryophytes.

Dehydration and rehydration of sperm cell suspensions of C. purpureus

To assess sperm dehydration tolerance after a single dehydration event and determine whether relative humidity (RH) during dehydration influences sperm recovery, we allowed sperm suspensions (5 μl each, placed within 0.6 ml open Eppendorf tubes) to dehydrate in airtight chambers. C. purpureus sperm was collected from 10 individuals of four populations. Suspensions were poised over one of four saturated salt solutions following protocols outlined by (Winston & Bates, 1960; Blackman et al., 1992), or over an anhydrous desiccant, (Drierite, W.A. Hammond Drierite Company Ltd, Xenia, OH, USA) for a minimum of 48 h. The saturated solutions used included: NaCl (75% RH), MgSO4 (55% RH), MgNO3 (33% RH), and anhydrous CaSO4. (< 10% RH). By 24 h, all samples were found to be visibly dry. To verify the stability of each RH treatment, temperature and RH within each chamber were monitored with micro data loggers (HOBO Pro v2; Onset Computer Corporation, Bourne, MA, USA). Average RH was as expected per treatment, and average temperature (± SD) was 21.15 ± 0.73°C. For rehydration treatments, at 48 h, 5 μl rainwater (at room temperature) was added to each tube, centrifuged for < 10 s at 3000 g, and left to rehydrate for a minimum of 2 h, at which time sample suspensions were gently vortexed and assessed for recovery using Cell Vu counting chambers as already described.

Addition of exogenous sugars to sperm cell suspensions of C. purpureus

In a subset of experiments, we added sucrose to sperm suspensions of C. purpureus to determine how the addition of exogenous sucrose influences sperm cell recovery from environmental desiccation stress. We followed the procedure outlined earlier with the addition of sucrose (0–25 mM final concentration) to the dehydration suspension or the rehydration suspension, or both, using C. purpureus sperm collected from six individuals from three populations. After 48 h (sucrose addition experiment) or 360 h (concentration experiment) of dehydration, the sperm samples were rehydrated and cells were assessed using Cell Vu counting chambers as already described.

Comparison of sperm cell tolerance with environmental desiccation among individuals of C. purpureus

To determine the amount of variation in tolerance among individuals to environmental desiccation, we extracted sperm suspensions (as already described) from nine individuals of C. purpureus. All plants were grown from the juvenile stage to maturity in a common-garden glasshouse environment. Three perigonia from three different ramets (clonally produced plants) were used for each of the nine individuals and maintained as separate samples, resulting in three replicate samples per individual. Sperm cells were collected, dehydrated, hydrated, and analyzed as described earlier. Cells were assessed at 72 h using Cell Vu counting chambers as already described.

Comparison of sperm cell longevity and tolerance to environmental desiccation among three moss species

To compare longevity and tolerance to environmental desiccation among species, sperm cells were collected and maintained hydrated or dehydrated (a subset with 25 mM sucrose-rainwater as the dehydration media), then rehydrated and analyzed as described earlier from glasshouse-grown individuals of B. argenteum, C. introflexus and C. purpureus, as well as field-collected plants of C. introflexus. Sperm cells were assessed in rainwater (fully hydrated) over the course of 96 h (4 d). In total, we examined sperm cell longevity in 60 samples of B. argenteum from three populations, 99 samples of C. purpureus from three populations, and 131 samples of C. introflexus. Owing to the clonal growth of mosses, we are not able to determine the number of individuals from our C. introflexus populations, but sampled broadly across populations.

Statistical analyses

Because the variance among treatments differed, we used a one-way Welch’s ANOVA to determine whether the number of intact C. purpureus sperm was affected by humidity treatment (< 10, 33, 55 and 75% RH; Morton & Forsythe, 1974). We used a simple one-way ANOVA to determine whether the number of intact C. purpureus sperm was affected by sucrose treatment (0.0, 0.25, 1.0, 10.0, and 25.0 mM). For a comparison of tolerance to dehydration among individual C. purpureus plants grown under common garden conditions, we used a mixed-model, nested ANOVA to determine whether recovery of sperm cells from dehydration was affected by population (random effect) or by individual (random effect, nested in population). For a comparison of tolerance to dehydration among species grown under common conditions, we used a mixed-model ANOVA to determine how species (random effect), treatment (fixed effect; hydrated, dehydrated, and dehydrated with sugar), time (fixed effect; 24, 48, 96 h), and interactions among these factors affected sperm counts. The original cell count was included as a covariate. For significant factors in this analysis, we used post-hoc analyses to determine which levels of that factor were significantly different from one another. All analyses were performed with JMP 8 (SAS Institute Inc., Cary, NC, USA, 2009).


Dehydration and ambient humidity

A fraction of the initial number of intact C. purpureus sperm cells were still intact after dehydration–rehydration events and were therefore deemed tolerant to environmental desiccation (17 ± 0.02%, mean for all treatments at 48 h; Fig. 1). Sperm was not significantly affected by humidity treatments (Fig. 1, F3,9.2 = 0.22; P = 0.88). The treatment with the lowest RH (< 10%, created using Drierite) showed the least variation among replicates.

Figure 1.

Effects of dry-down speed on sperm cell recovery from environmental desiccation. Fraction of Ceratodon purpureus sperm intact after 48 h in four dehydration–rehydration treatments with variable relative humidity (n = 30). Error bars represent data means ± SE. Means do not vary significantly among treatments.

Dehydration and sucrose addition

The addition of sucrose to the suspension media significantly affected cell recovery of sperm in dehydration treatments of C. purpureus, but timing of the addition was critical (Fig. 2a). We found that after 48 h of environmental desiccation, three treatments varied significantly (F = 13.17; P = 0.0002). The samples with sucrose added to the dry-down media showed significantly higher recovery from dehydration than the samples without sucrose added at either time (P = 0.0003), and from those with 25 mM sucrose added to the rehydration media only (P = 0.0019). The sperm samples with sugar added to the rehydration media only showed slightly higher recovery than those without any sucrose added, but the results were not significant. We found that the concentration of sucrose at dehydration significantly affects sperm cell recovery after an extended period of dehydration (360 h; F = 7.54; P = 0.0002, Fig. 2b). Specifically, Tukey post-hoc analysis indicates that 25 mM differed significantly from the control, 0.0 mM (P = 0.0003), 0.25 mM (P = 0.0009), and 1.0 mM (P = 0.02), and notably was the only concentration that differed significantly from the control or from any other pairwise combination.

Figure 2.

Exogenous sucrose aids sperm cells in recovery from environmental desiccation events. Fraction of Ceratodon purpureus sperm cells intact after dehydrationrehydration treatment with or without the addition of sucrose. (a) Fraction of sperm intact after 48 h in one of three treatments: −/−, rainwater for dehydration and rehydration; −/+, rainwater for dehydration, sucrose was added to rehydration media; +/−, sucrose was added to dehydration media, rainwater for rehydration (n = 24, P < 0.0003). Bars represent data mean, whiskers are minimum to maximum values. Different letters represent means that differ significantly from one another (P < 0.005). (b) Fraction of sperm intact after 360 h of dehydration–rehydration treatment with increasing concentrations of sucrose added to the dehydration media (n = 36; P = 0.0002). Error bars represent data means ± SE.

Comparison among individuals

Individuals of C. purpureus demonstrated variance in the fraction of their sperm that are tolerant to environmental desiccation (Fig. 3), although all individuals had desiccation-tolerant sperm. Individuals showed significantly different recovery from environmental desiccation (F = 8.53; P = 0.006). Original cell number did not significantly affect the model (F = 2.72; P = 0.14) and populations did not vary significantly in recovery (F = 0.70; P = 0.63).

Figure 3.

Variation among individuals in sperm cell recovery from environmental desiccation. Fraction of Ceratodon purpureus sperm intact after 72 h dehydration–rehydration treatment from nine individuals of four populations. (n = 18, P = 0.006). Error bars represent data means ± SE.

Comparison among species

Sperm cell recovery was not significantly different among species (n = 277; Fig. 4; Table 1). Sperm cell recovery was significantly affected by treatment and time. Cell recovery was significantly higher in sperm that remained hydrated than in those that were dehydrated with and without sucrose (Fig. 4). The three species of moss sperm, when hydrated, maintained a high percentage of intact cells for an extended period of time, in the range of d rather than h (Fig. 4a). Cell counts decreased significantly with each increasing time point. Sperm that were dehydrated with sugar had significantly higher survival than those that were dehydrated without sugar (Fig. 4b,c). The interaction between treatment and time was also significant, as survival declined rapidly in hydrated treatment, but was not significant in either the dehydrated treatment or the dehydrated treatment with sugar (Fig. 4). A three-way interaction of treatment, time and species was not significant.

Figure 4.

Effects of time and dehydration treatments on sperm cell integrity in three species of mosses. Closed circles, Bryum argenteum; open circles, Ceratodon purpureus; squares, Campylopus introflexus. (a) Hydration treatment: fraction of sperm still intact over time. The sperm were stored in rainwater and assayed at 24, 48 and 96 h after sperm extraction (n = 109). Error bars represent data means ± SE. (b) Dehydration treatment: fraction of sperm still intact after dehydration–rehydration events. The sperm was rehydrated at 24, 48, and 96 h and assessed for cellular integrity (n = 125). Error bars represent data means ± SE. (c) Sucrose and dehydration treatment: fraction of sperm cells intact after dehydration–rehydration events with sucrose added to dehydration media The sperm was rehydrated at 24, 48, and 96 h and assessed for cellular integrity (n = 56). Error bars represent ± SE.

Table 1.   Results of a mixed-model ANOVA (n = 277) to determine how species (random effect; Bryum argenteum, Campylopus introflexus, and Ceratodon purpureus), treatment (fixed effect; hydrated, dehydrated or dehydrated with sucrose), time (fixed effect; 24, 48, 96 h), and interactions among these factors affected sperm cell recovery
 dfSS F P
  1. The original cell count before treatment was included as a covariate. Significant P-values are indicated in bold. SS, sum of squares.

Treatment16.72181.48 < 0.0001
Species × time30.272.250.17
Species × treatment30.161.290.37
Time × treatment31.238.19 0.003
Species × time × treatment70.281.340.23
Original cell count10.0010.050.83

In a subset of treatments of B. argenteum and C. introflexus sperm, sucrose was included in the rainwater media and they were assessed for longevity at 48 h. Interestingly, the addition of sucrose to the media significantly affected cell longevity (= 29, F = 4.65, P = 0.01); at 48 h, the mean (± SE) recovered cell count of sperm suspended with added sucrose was lower (51.2 ± 0.004%) than sperm without sucrose added to the hydration media (62.0 ± 0.004%), suggesting that sperm cells do not utilize sucrose to increase longevity.

In all cases, we found that only a fraction of sperm from any individual survived environmental desiccation, suggesting variation within individuals. This result indicates that there may be different types of sperm within a single individual. Additionally, in all three species tested there appeared be a bimodal distribution in sperm motility within individual antheridia (E.E. Shortlidge, T.N. Rosenstiel, & S.M. Eppley, unpublished). Generally, our observations suggest that slower sperm are more tolerant to environmental desiccation, but robust correlation will require further work.


Although the strategy of freely releasing sperm into the environment is widespread and found in a diverse array of taxa, little attention has been given to determining how motile sperm maintain viability and function in the environment. Here, we examined the ability of moss sperm to tolerate extreme dehydration, such as they might encounter in the terrestrial landscape. We found that sperm from three moss species, from three families, display a significant capacity to tolerate environmental desiccation. All three mosses produce sperm cells capable of maintaining integrity over an extended period of time when hydrated (Fig. 4a), and of recovering from environmental desiccation events (Fig. 4b). To our knowledge, this study represents the first to explore tolerance to environmental desiccation in free motile sperm cells. Our results suggest that moss sperm may have the potential to persist for extended periods of time on the landscape. Such findings could have significant implications for our understanding of moss mating systems and the factors that influence moss reproductive success, including the possibility of a sperm bank existing on the landscape, increased opportunity for fertilization, and a new trajectory for scientific investigation into sperm transport. In the following, we discuss the characteristics of tolerance to environmental desiccation in moss sperm with respect to two characteristics that are common in desiccation tolerance in plants (i.e. correlation with dry-down rate and increased tolerance with addition of sugars), and variation in measured traits among moss sperm.

Tolerance to environmental desiccation

For most organisms, rate of dehydration correlates with recovery from desiccation events, with a slower rate of dehydration generally resulting in increased recovery upon rehydration (e.g. Schonbek & Bewley, 1981; Womersley & Ching, 1989). For instance, vegetative tissues of the desiccation-tolerant moss Tortula ruralis resumed RNA synthesis quickly upon rehydration when the desiccation rate was slow vs fast, as a result of adequate time for up-regulation of cellular water-loss responses (Oliver & Bewley, 1984; Oliver et al., 2000a; Proctor et al., 2007). However, in reproductive plant tissues such as seeds, in which desiccation is a programmed event during maturation, desiccation tolerance increases with faster dry-down, resulting in higher survival, presumably because less time is spent at intermediate water levels, where more damage can occur (Pammenter & Berjak, 1999; for review). Our data show that the RH at which moss sperm dry-down is not correlated to their recovery. This result indicates that the rate of dry-down did not affect recovery rate. We did find that the lowest RH, and therefore the fastest dry-down speed, yielded the least variance around the mean for cell integrity. We know that vegetative desiccation-tolerant plants that can survive rapid dry-down rely on both cellular repair and cellular protective mechanisms (Oliver et al., 2000b). Whether moss sperm are able to undergo an up-regulated stress response for cellular repair or rely only on constitutive protective mechanisms remains uncertain, but our results of similar sperm cell recovery among humidity treatments (Fig. 1) indicate that the mechanism may be constitutive.

While we do not see a correlation with respect to rate of dry-down and recovery in our data, our results support the role of sugars in facilitating tolerance to extreme dehydration. We found that the fraction of cells recovering from desiccation was significantly increased by the addition of exogenous sucrose at dry-down (Figs 2a, 4c), and increased as a factor of increasing exogenous sucrose concentration (Fig. 2b). An extensive body of literature demonstrates the role of nonreducing disaccharides in protection from desiccation damage in prokaryotes and eukaryotes (e.g. Crowe et al., 1984, 1992; Potts, 1994; Ingram & Bartels, 1996; Hoekstra et al., 2001; Garg et al., 2002). Studies have also examined the role of sugars after desiccation events in plant vegetative structures (e.g. Bewley et al., 1978; Bianchi et al., 1991; Smirnoff, 1992; Garg et al., 2002), seeds (e.g. Koster & Leopold, 1988; Koster, 1991; Sun et al., 1994), and pollen (e.g. Hoekstra & Vanroekel, 1988). While no other studies have, to our knowledge, examined whether sugars affect stress tolerance in plant sperm, the impact of sugars on mouse sperm has been examined. Two studies found that adding exogenous trehalose to mouse sperm increased tolerance to desiccation (Bhowmick et al., 2003; McGinnis et al., 2005). The researchers speculated that increased sperm viability was a result of the relative ease with which intracellular osmolytes could come to equilibrium with a dry-down media fortified with trehalose. A similar phenomenon may have occurred in our system.

The role of sucrose in desiccation tolerance of moss sperm is only ecologically relevant if sperm encounter sucrose in the environment. Interestingly, sucrose may be present in the terrestrial environment in which moss sperm are released. Not only does sucrose accumulate inside moss cells that endure desiccation and freezing events (Rütten & Santarius, 1992; Proctor et al., 2007), but sucrose is a major component of the pulse-released suite of compounds found after a rewetting event, post-desiccation (Coxson et al., 1992). Furthermore, sucrose is a constituent of the exudates from moss archegonia (Pfeffer, 1884; Kaiser et al., 1985; Ziegler et al., 1988) and potentially of other vegetation and fruits found in the plant canopy above. Hence, there is evidence of a multifaceted role for sucrose in the ecophysiology of mosses.

Sperm variation

While we speculate that an individual moss will be at an advantage if its sperm tolerates environmental desiccation, genetic variation must be present for selection to act on the trait. Genetic variation for desiccation tolerance among populations and individuals has been found in many organisms, including plants (e.g. Glazer et al., 1991; Basnayake et al., 1993; Hoffmann & Harshman, 1999; Jurenka et al., 2007). In Drosophila melanogaster, Hoffmann & Parsons (1989a) found that desiccation resistance is up to 60% heritable. High amounts of genetic variation suggest this trait would respond rapidly to selection, which has been confirmed by experiments (e.g. Chippindale et al., 1998). We found that variation in sperm tolerance to environmental desiccation does not occur among examined species, but does occur among individuals within a species (Table 1; Fig. 3). Our mosses were grown in common-garden, uniform glasshouse conditions (although they were collected from various locales), and therefore differences among individuals are likely genetic, although we cannot rule out increased variance in maternal effects, as we collected individuals from within and among populations. However, the most likely explanation is that genetic variation exists among individuals for sperm tolerance to environmental desiccation, leading to the question as to why such a potentially adaptive trait would not sweep through a population. Researchers have suggested that an organism’s ability to tolerate any stress, not just desiccation, is inversely proportional to its ability to compete because of a tradeoff between stress tolerance and growth or metabolic rate (Grime, 1977; Hoffmann & Parsons, 1989; Chapin et al., 1993; Arendt, 1997). Research on echinoderms implies tradeoffs in sperm velocity vs life span (Levitan, 2000), suggesting that energetically based life history tradeoffs can occur in sperm. Further research will be essential to determine whether there are tradeoffs between differential stress tolerance in moss sperm and other sperm traits, and how environmental conditions may inform these relationships.

We also found that only a fraction of sperm from any individual of our three species survived desiccation, suggesting variation within individuals among sperm in tolerance to environmental desiccation, and perhaps suggesting the possibility of variable sperm types within a single moss individual. Heteromorphic sperm are widespread among animal taxa, including in such unrelated lineages as molluscs, insects, echinoderms, and vertebrates (e.g. Lee & Wilkes, 1965; Hodgson, 1997; Au et al., 1998; Buckland-Nicks et al., 1999; Lutzen et al., 2004; Hayakawa et al., 2007; Van Look et al., 2007), many of which reproduce via external fertilization. Heteromorphic sperm also occur in angiosperms (e.g. Saito et al., 2002; Weterings & Russell, 2004; Gou et al., 2009; Hirano & Hoshino, 2010). Heteromorphic pollen in angiosperms, and sperm in animals have been suggested to be convergent, having both evolved via sexual selection to increase male success (Till-Bottraud et al., 2005). Sperm with variable DNA content have been documented within individual bryophytes (Renzaglia et al., 1995), and researchers have speculated that, because of the prevalence of heterospermy across animal taxa, variable sperm should also be present within nonseed plants (Till-Bottraud et al., 2005). Although the moss sperm are formed through a mitotic pathway, there is potential for variation in spermatogenesis, such as during cytoplasmic reduction, or as a result of environmental treatments, as found in some ferns (Southworth & Cresti, 1997; for review). The ecological significance of heteromorphic sperm is unknown in angiosperms (Weterings & Russell, 2004). However, in animals, which (like moss) also release naked sperm, variation in sperm type within an individual is widespread. Among animals, heteromorphic sperm can vary in DNA content and fertilization ability, and have been hypothesized to have important co-functions in reproduction, for example, protecting co-sperm from harsh environmental conditions, aiding in sperm competition, and facilitating movement of co-sperm in aqueous environments (Silberglied et al., 1984; Buckland-Nicks, 1998; Hayakawa, 2007; Higginson & Pitnick, 2011). Whether the moss sperm in our study are truly heteromorphic with different developmental pathways, constitutive defenses, DNA contents, and/or abilities to fertilize, or whether they vary in tolerance to desiccation across an environmental spectrum remains to be seen. Future studies should not only consider the possibility of heteromorphic moss sperm and the implications for the evidently complex reproductive ecology of mosses, but also seek to understand the fitness consequences for individuals possessing functional variation in sperm traits.


We thank A. Dittmer, T. Hinkley, and M. Slate for help in the glasshouse, and H. Le, C. Rupert, and J. Storey for laboratory support. Research was supported by the US National Science Foundation (DEB 0743461 to S.M.E. and IOS 0719570 to T.N.R.) and 3M Corporation.