Species-specific effects of passive warming in an Antarctic moss system

Polar systems are experiencing rapid climate change and the high sensitivity of these Arctic and Antarctic ecosystems make them especially vulnerable to accelerated ecological transformation. In Antarctica, warming results in a mosaic of ice-free terrestrial habitats dominated by a diverse assemblage of cryptogamic plants (i.e. mosses and lichens). Although these plants provide key habitat for a wide array of microorganisms and invertebrates, we have little understanding of the interaction between trophic levels in this terrestrial ecosystem and whether there are functional effects of plant species on higher trophic levels that may alter with warming. Here, we used open top chambers on Fildes Peninsula, King George Island, Antarctica, to examine the effects of passive warming and moss species on the abiotic environment and ultimately on higher trophic levels. For the dominant mosses, Polytrichastrum alpinum and Sanionia georgicouncinata, we found species-specific effects on the abiotic environment, including moss canopy temperature and soil moisture. In addition, we found distinct shifts in sexual expression in P. alpinum plants under warming compared to mosses without warming, and invertebrate communities in this moss species were strongly correlated with plant reproduction. Mosses under warming had substantially larger total invertebrate communities, and some invertebrate taxa were influenced differentially by moss species. However, warmed moss plants showed lower fungal biomass than control moss plants, and fungal biomass differed between moss species. Our results indicate that continued warming may impact the reproductive output of Antarctic moss species, potentially altering terrestrial ecosystems dynamics from the bottom up. Understanding these effects requires clarifying the foundational, mechanistic role that individual plant species play in mediating complex interactions in Antarctica's terrestrial food webs.

Polar systems are experiencing rapid climate change and the high sensitivity of these Arctic and Antarctic ecosystems make them especially vulnerable to accelerated ecological transformation. In Antarctica, warming results in a mosaic of ice-free terrestrial habitats dominated by a diverse assemblage of cryptogamic plants (i.e. mosses and lichens). Although these plants provide key habitat for a wide array of microorganisms and invertebrates, we have little understanding of the interaction between trophic levels in this terrestrial ecosystem and whether there are functional effects of plant species on higher trophic levels that may alter with warming. Here, we used open top chambers on Fildes Peninsula, King George Island, Antarctica, to examine the effects of passive warming and moss species on the abiotic environment and ultimately on higher trophic levels. For the dominant mosses, Polytrichastrum alpinum and Sanionia georgicouncinata, we found species-specific effects on the abiotic environment, including moss canopy temperature and soil moisture. In addition, we found distinct shifts in sexual expression in P. alpinum plants under warming compared to mosses without warming, and invertebrate communities in this moss species were strongly correlated with plant reproduction. Mosses under warming had substantially larger total invertebrate communities, and some invertebrate taxa were influenced differentially by moss species. However, warmed moss plants showed lower fungal biomass than control moss plants, and fungal biomass differed between moss species. Our results indicate that continued warming may Antarctic Peninsula is anticipated to encourage the growth and spreading of mosses in Antarctica, which will further influence terrestrial ecosystem succession [8,25,36] and hold implications for the microorganism and invertebrate communities they sustain [8]. Moss growth can be clonal or sexual, and only a few studies on Antarctic mosses have examined the effects of warming on sexual reproduction [35,37], which will be key to understand changing moss communities. Furthermore, despite the important ecological connection of invertebrates and moss fertilization [38,39], no studies to date have examined both reproductive expression and associated invertebrate communities of the dominant moss species.
Given the magnitude of moss cover in ice-free maritime Antarctica, it is crucial to study speciesspecific responses to warming and how they may link to terrestrialization processes in the Antarctic. Here, we tested the effects of increased temperature on Antarctic moss communities using open top chambers (OTCs) established for 8 years (2008-2016) to simulate warming and control plots (unwarmed) at Juan Carlos Point on Fildes Peninsula, located in maritime Antarctica on King George Island (KGI). In the experiment, we examined vegetation species cover and collected cores of the two dominant moss species, Polytrichastrum alpinum and Sanionia georgicouncinata, which comprise 65% of the terrestrial vegetation cover in the area. By using these cores, we tested the effects of warming on moss canopy morphology, moss reproductive effort and moss-associated invertebrate communities and fungal biomass. We predicted that passive warming would differentially impact Antarctic moss species, highlighting underlying moss species-specific differences and that these differences would further scale to affect the Antarctic terrestrial moss-associated food web.

Study site and species
Our study site was located in the South Shetland Island archipelago on Fildes Peninsula, KGI, one of the largest ice-free areas in the region [17]. KGI is known for its rich bryophyte flora, containing 61 bryophyte species, of which 40 have been recorded on Fildes Peninsula [17]. Our warming experiment was carried out at Juan Carlos Point (62°12 0 S, 58°59 0 W), a moist coastal lowland site with northern exposure towards Drake Passage and a rich moss-grass community, a community type commonly found on several islands along the South Shetland Archipelago [37,40]. This community is dominated by the grass Deschampsia antarctica (Desv.) and two moss species, P. alpinum (Hedw.) and S. georgicouncinata (Hedw.). These two moss species have bipolar distributions and are widespread and common species in this region of Antarctica [17]. Polytrichastrum alpinum is dioecious, with separate male and female plants, while S. georgicouncinata is hermaphroditic, and both species are rarely sexually reproductive in Antarctica, with only a single record of plants with sporophytes for S. georgicouncinata [17]. The previous research from this same warming experiment demonstrated that P. alpinum increased sporophyte production with warming [37]. Similarly, a warming experiment at a nearby site found that gametangia increased significantly in P. alpinum plants in warmed versus control plots [35].

Passive warming experiment
As a part of the long-term warming experiment, we established 10 OTCs and matched controls in 2008 at Juan Carlos Point [37] (electronic supplementary material). OTCs are designed to produce an increase in air temperature by preventing the loss of heat by convection processes and have been commonly used in both Arctic and Antarctic ecosystem studies [41,42]. The OTCs used here are similar to those used in other warming studies located in Antarctica [35,37,43]. The chambers are composed of 3 mm thick, transparent acrylic panels, assembled into hexagonally shapes that taper to an open top of 40 cm height and a basal footprint of 106.4 cm 2 . The acrylic walls have small perforations to allow air exchange and avoid excessive warming. The OTCs were placed on areas identified by researchers as having approximately 80-90% plant cover (with moss cover approx. 50% and the remaining percentage being lichen), and OTCs remained in place year-round. Matched control sites were marked within 1 m of each OTC and were selected with similar plant composition and cover, but did not receive treatment [37].
Previously, we have shown that the OTC treatment in this experiment significantly increased mean maximum daily air temperature during the study period (2008-2010), from 7.3°C in control plots to 10.5°C in OTCs [37]. Temperature and humidity data were collected using HOBO Pro v2 loggers (Onset Computer Corporation, Bourne, MA). However, the passive warming treatment had no significant effect on mean daily temperature or mean minimum daily air temperature (analysed annually or including royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 190744 season as a factor; mean site minimum daily air temperature at Juan Carlos Point, −6.4°C). The highest warming effect measured was during the summer season, with an increase of 0.61°C inside the OTCs compared to control plots. These values are similar to other reported values for passive warming experiments in maritime Antarctica, where Bokhorst et al. [44] measured an increase of 0.7°C in annual mean temperature inside OTCs when compared to control plots. The use of OTCs also produced changes in microclimate as mean daily relative humidity was significantly lower in the OTCs (80.7%) than in control plots (91.7%) [37].

Moss species effects on abiotic conditions
To determine the impact of moss species identity on the microenvironment temperature within the warming experiment, moss canopy temperatures were measured during the austral summer of 2015. Moss canopy temperatures were measured in contiguous patches of each dominant moss species, P. alpinum and S. georgicouncinata, using thermocouple data loggers inserted approximately 0.5 cm into the moss canopy (HOBO, UX120-014M, Onset Computer Corporation, Bourne, MA), in both an OTC and a control plot. Recordings were made every 10 s from 9 to 18 January 2015.
To assess how moss species influence soil temperature and moisture, we constructed separate arrays of P. alpinum and S. georgicouncinata. Arrays consisted of 8 × 13 cm blocks of each moss species (with each block being the depth of the moss to the rhizoid level) paired with a control of bare soil. Soil temperature (°C) and volumetric water content (m 3 m −3 ) were measured once an hour below moss species (2 cm below soil surface, below the rhizoid level) and in bare ground control plots (2 cm below soil surface) during the austral summer season in 2016 using Decagon EM50 and 5TM soil moisture and temperature sensors (Meter Group, Inc., Pullman, WA).

Plant and cryptogam community analysis
Plant and cryptogam communities were assessed by sight for the per cent cover in the summer season 2015, for seven pairs of OTC and control plots, using a 0.5 m square frame placed in permanently marked blocks within plots. Two researchers conducted the surveys, and all points were agreed upon to eliminate single observer bias. Lichens were grouped to genus level, and crustose forms were categorized by observable morphology. One vascular grass species, D. antarctica, was recorded in the experimental plots.

Moss canopy morphology
We quantified eight moss canopy characteristics by extracting 2 cm diameter cores of P. alpinum and S. georgicouncinata from each of five matched OTC and control plots in 2016. Intact cores, which included the entire moss profile from rhizoid to gametophyte top, were weighed to the nearest milligram (dry weight (DW)), and electronic calipers were used to measure the height of the intact core on randomly selected locations. Tissue height measurements (to the nearest millimetre) were conducted in four categories: photosynthetic tissue height, senescent tissue height, soil/rhizoid height and total core height. In addition, the height of five randomly selected gametophytes from each core was measured to the nearest millimetre to estimate the canopy height. Mean canopy height was calculated from the gametophyte measurements, and mean core height was calculated from the intact core measurements. Canopy density was determined by counting the number of gametophytes per core; leaves were denuded from one randomly selected gametophyte from each core to determine leaf density.

Reproductive biology and expressed sex ratios in Polytrichastrum alpinum
Polytrichastrum alpinum cores (2 cm diameter) from five matched OTC and control plots were separated, and each gametophyte was counted, confirmed as P. alpinum and categorized by age as juvenile (very green with leaves still appressed), mature (leafy gametophyte) or senescent (very little to no green tissue, blackening), following the methods of Shortlidge et al. [35], which are modified from a standard bryophyte protocol for placing leaves in age classes using colour and morphology [45]. Ten randomly selected gametophytes from the mature age category were dissected and examined for the presence of sex expression under compound and dissecting microscopes (Leica Application Suite 3.5.0, Leica, Germany). Examination of plants under the microscope is the standard protocol for determining sex expression for plants in this family [46]. We classified sex expression based on the presence or absence of male sex organs (antheridia), female sex organs (archegonia), sporophytes or paraphyses (sterile reproductive structure). If gametophytes had no royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 190744 sexual expression, they were classified as purely vegetative (no evidence of gametangia, sporophytes or paraphyses). Laboratory determinations of sexually expressing gametophytes per core were combined and averaged for OTC and control plots.

Bryophyte-associated invertebrate communities
We quantified the invertebrate community ( presence, absence and abundance) from the dominant moss species (P. alpinum and S. georgicouncinata) by extracting replicate 2 cm diameter cores from paired OTC and control plots, yielding 23 cores.
Invertebrates were extracted using an innovative dried substrate extraction technique, which was tested and found to yield comparable extraction numbers as matched extractions performed with Tullgren extractors from the same site location. Moss cores were placed directly into paper coin envelopes and dried in an oven at 30°C for 72 h. Each core was then carefully removed, placed into a weigh boat and core mass quantified to the nearest milligram. Cores were homogenized carefully by hand and divided into two 50 ml Falcon tubes. Invertebrates were extracted by vigorously shaking the tubes using a blend of kerosene and 95% ethanol (1 : 3 ratio). The solution was allowed to settle, and transfer pipettes were used to extract invertebrates from the kerosene suspension into a Petri plate for identification under dissecting microscope. Extracted invertebrates were sorted taxonomically into the following groups: Oribatida, Collembola and Nematoda. Invertebrate abundance was expressed as number of individuals per sampled moss core.

Estimation of bryophyte fungal biomass
Dried 2 cm diameter cores of P. alpinum and S. georgicouncinata were separated from the soil and lyophilized. For each sample, 100-200 mg of lyophilized tissue was placed in a 2 ml locking microcentrifuge tube with a zirconium bead and ground to a fine powder using a bead beater at 2500 r.p.m. for 40 s. Ergosterol was extracted from the powdered tissue following the methods of Dahlman et al. [47], with the following changes. Samples were suspended in 1 ml MeOH, agitated in an orbital shaker for 1 h at 320 r.p.m. in darkness and subsequently allowed to precipitate overnight at 4°C. Extractions were then centrifuged for 1 h at 14 000 r.p.m. and 4°C. Following centrifugation, the supernatant from each extraction was transferred to a fresh 2 ml microcentrifuge tube and centrifuged for an additional 10 min at 140 000 r.p.m. and 4°C. The supernatants from this secondary centrifugation were filtered through 0.2 µm acrodiscs into amber autoanalyser vials and stored at 4°C until high-performance liquid chromatography (HPLC) analysis. Extractions were analysed following [48] on a 1200 Series HPLC (Agilent Technologies, Waldbronn, Germany).

Statistical analyses
We used two-way analysis of variance (ANOVA) to determine the effect of treatment (OTC versus control), moss species (P. alpinum versus S. georgicouncinata) and the interaction between these factors on canopy temperature. We used ANOVA to determine the effects of ground cover type (P. alpinum versus S. georgicouncinata versus bare ground) on soil temperature and soil moisture. We employed a Tukey's honest significant difference (HSD) to examine pairwise contrasts among ground cover types. We analysed and visualized these data in the R statistical platform [49] version 3.3.3. with the packages: lubridate [50], ggplot2 [51] and dplyr [52].
To reveal potential changes in vegetation community assemblages between OTC and control plots, a per cent cover matrix was ordinated using non-metric multidimensional scaling (NMDS) in R studio [49] version 0.99.473 using package vegan [53]. By using Bray-Curtis distance measures and square root data transformation, we chose a two-dimensional solution with the best stress (0.09) based on 20 runs of 1000 permutations. An analysis of similarity (ANOSIM) was used to statistically test for differences between warmed and unwarmed cryptogam species assemblages.
We used a generalized linear model (GLM) to determine the effect of treatment (OTC versus control), bryophyte species (P. alpinum and S. georgicouncinata) and the interaction between these factors on core DW (with a normal distribution and an identity link function) and photosynthetic tissue height (with a Poisson distribution and a log link function) for the cores collected for each species for plant morphological analysis (separate bryophyte cores were collected for invertebrate analysis). For these analyses, we included subsite (the location of each paired OTC and control plot), but it was not significant and was dropped from the final analyses. We used t-tests to determine the effect of treatment on P. alpinum canopy density (number of gametophytes per core), number of juvenile royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 190744 gametophytes per core and gametophyte density (number of leaves per gametophyte), as these more intensive measures were only available for this species.
To determine the effect of treatment (OTC and control) and subsite (the location of each paired OTC and control plot) on gametophytes with female sex expression (archegonia) and gametophytes with sterile sex structures ( paraphyses) in P. alpinum, we used a GLM with Poisson distribution with a log link to account for the large number of zeros in the dataset, which represent low-frequency effects rather than missing data values. No gametophytes with male sex expression were found.
We used simple regression analysis to determine whether nematode or microarthropod (Collembola and Oribatida) abundance was influenced by plant biomass. We used GLMs to test the effect of moss species (P. alpinum and S. georgicouncinata), treatment (OTC versus control), the interaction between these factors, subsite on the number of nematodes per bryophyte core (using a normal distribution and an identity link function) and the number of microarthropods per bryophyte core (using a Poisson distribution and log link function). We used a simple regression to determine whether microarthropod abundance was correlated with female sexual expression. We used a GLM to test the effect of moss species (P. alpinum and S. georgicouncinata), treatment (OTC versus control) and the interaction between these factors on ergosterol content, used as a proxy for bryophyte fungal biomass. All analyses were conducted using JMP version 14.1.0 [54], unless otherwise stated.

Moss effects on abiotic conditions
Temperatures in both P. alpinum and S. georgicouncinata mosses were significantly higher in OTC compared to control plots (n = 172,800, F = 9906.59, p < 0.0001, figure 1a). While S. georgicouncinata   control plots were significantly warmer than P. alpinum controls (n = 86 400, F = 5.51, p = 0.02, figure 1a), the opposite was true for OTC treatments. The interaction of treatment and species was also significant (n = 86 400, F = 643.73, p < 0.0001, figure 1a). Bare soils were significantly warmer than soils under either P. alpinum or S. georgicouncinata, in addition, moss identity did not significantly impact soil temperatures (n = 744, F = 68.35, p < 0.0001, figure 1b). Soil moisture content was drastically lower in bare soil treatments than in soil under either moss species (n = 744, F = 32 546.00, p < 0.0001, figure 1c). Polytrichastrum alpinum cover provided a significant increase in soil moisture when compared with soils covered by S. georgicouncinata or in bare soil (n = 744, F = 32 546.00, p < 0.0001, figure 1c).

Cryptogam community response to passive warming
Across the seven OTC and control plots surveyed, we found seven bryophyte species, Bartramia patens, Bryum pseudotriquetrum, Meesia uliginosa, Pohlia sp., P. alpinum, S. georgicouncinata and Syntrichia saxicola. Two species of bryophytes occurred only in OTC plots, B. patens and Pohlia sp. In addition, NMDS ordination of cryptogam communities showed that community composition showed no detectable compositional changes (as measured by the per cent cover) after 8 years of passive warming (n = 7, NMDS stress = 0.09, ANOSIM test statistic p = 0.35). All plots were abundant with vegetation, and no bare ground or rock substrate was recorded, which is as expected given the site lies in a moist, coastal lowland that is rich with cryptogamic vegetation.

Sex expression in Polytrichastrum alpinum
In P. alpinum, plants in OTCs produced female reproductive structures (archegonia) in significantly greater numbers than did plants in the control plots (d.f. = 1, X 2 = 5.65, p = 0.02, figure 2). Subsite (the location of the paired OTC and control plots) was significant in this analysis (d.f. = 4, X 2 = 14.56, p = 0.006). The effect of the warming treatment on the occurrence of paraphyses (sterile reproductive structures) was not significant (d.f. = 1, X 2 = 3.80, p = 0.051). No male sex expression (antheridia) or sporophytes were observed during the sampling season.

Bryophyte fungal biomass under passive warming
We used ergosterol content as a proxy for fungal biomass in two dominant moss species and found that moss species significantly affected the bryophyte fungal biomass. Extracted cores of P. alpinum had significantly more fungal biomass than S. georgicouncinata (n = 12, d.f. = 1; X 2 = 20.00; p < 0.0001,

Discussion
Here, we present results of 8 years of experimental warming on a moss ecosystem in maritime Antarctica. Experimental warming had distinct and significant impacts on both microclimate conditions and bryophyte communities. We found that our OTC warming treatments were successful at significantly increasing air temperature during the study period, as they are designed to do, and in addition, moss canopy temperatures also significantly increased in OTC treatments. Relative humidity, as measured in air, significantly decreased in OTCs. Our abiotic results are similar to those generated in other Antarctic studies utilizing OTC chambers; see the study by Bokhorst et al. [42] for review on passive warming methods in the Antarctic. Yet, a large variation in abiotic conditions has been generated using OTCs in Antarctica, some of which are dissimilar to our findings. This variation demonstrates the variance in terrestrial, ice-free sites found in Antarctica and the varied methods of measurement (air, soil and moss canopy). Microsite differences are expected to be critical in explaining cryptogam survival during climate change in Antarctica, and more detailed understanding of topography and plant health is needed to understand this relationship [55][56][57].
Results from this passive warming experiment demonstrate that moss species were affected differently by warming and in turn differentially affected communities at the higher trophic level. Our results indicate that increasing warming in the Western Antarctic Peninsula will have distinct impacts on Antarctic moss sexual expression, and species-specific effects of these dominant land plants may hold implications for the terrestrial community of microorganisms and invertebrates that these mosses sustain. We discuss these results below.

Species-specific thermal effects observed in bryophytes under experimental warming
We found that mosses in OTCs had significantly higher canopy temperatures than those located in control plots, and significant species-specific canopy temperature differences also occurred, with higher canopy temperatures occurring in P. alpinum ( figure 1a). These species-specific differences in canopy temperatures are probably due to albedo effects, as has been found in subarctic cryptogams [30]. Studies in higher latitudes suggest that moss species differ not only in their ability to absorb heat but also in their ability to transfer thermal energy to the soil [29]. We found that bare soils were significantly  . Mean (±s.e.) ergosterol content (µg g −1 DW moss tissue) extracted from two dominant moss species (P. alpinum and S. georgicouncinata) in warmed (OTC) and control (CON) plots. Ergosterol content was significantly different among treatments (T) and P. alpinum had significantly more fungal biomass than S. georgicouncinata (S). The interaction between treatment and species (SxT) was not significant (n = 12).
royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 190744 warmer than when moss cover was present (figure 1b), but that moss species were similar in their insulating effects. Moss species did significantly affect soil moisture; P. alpinum provided a significant increase in soil moisture when compared with soils covered by S. georgicouncinata (figure 1c). Our results indicate that climate warming along the Western Antarctic Peninsula may alter the way that moss canopies influence energy exchange between the atmosphere and soils, with implications for the water-soil-atmosphere interface in a warming Antarctica. Moss cover can reduce the exchange of heat between the atmosphere and the soil, providing an insulating effect that decreases soil temperatures and protects permafrost [58,59] and further can provide habitat for microfauna adapted to the Antarctic ecosystem [60][61][62]. A reduction of surface cover of cryptogams or changes in the insulating and/or water-holding capacity of moss species may hold significant effects for future soil temperatures, soil moisture, thawing of permafrost, communities of microfauna and processes of ecosystem succession along the Western Antarctic Peninsula [23,25,58,63].

Effects of experimental warming on moss species
After 8 years of passive warming, we observed no distinguishable effects on cryptogam community composition overall, a finding consistent with other Antarctic passive warming studies [35,37,43,64]. However, we did find differences in sexual expression; P. alpinum in warmed (OTC) plots exhibited significantly more female sex expression than control plots. Our results are similar to those of Shortlidge et al. [35], who found increases in individual gametangia production and decreases in cellular stress defenses, suggesting that warming may relieve environmental constraints on reproductive expression. Our findings also identified a correlation between the number of female moss reproductive structures and the abundance of microarthropods within the moss canopy (r 2 = 0.64). Studies show that both springtails and mites prefer sexually expressing moss shoots rather than non-expressing shoots and that microarthropods prefer female over male shoots [38,39]. Microarthropods may prefer sexually expressing plants directly or indirectly (via fungi) because of sugars, starches and fatty acids excreted by gametangia [39,65,66]. Our study cannot determine whether microarthropods are increased on plants with gametangia because of these reproductive structures or because of an unmeasured factor that also correlates with sexual expression in our study. However, studies show that sexual reproduction in mosses is facilitated by springtails and mites [38,39], and because microarthropods may benefit more nutritionally from sexually expressing versus non-expressing plants [39,65,66], understanding the potential relationships among climate warming, sex expression in mosses and microarthropod and moss population dynamics will be instructive to understand the interaction between mosses and microarthropods in Antarctica.

Warming and moss species differentially affect higher trophic levels
Invertebrate communities extracted from bryophytes responded differentially under 8 years of passive warming treatment (OTCs) compared with controls. We found that overall microarthropod abundance increased in warmed (OTC) plots compared with control plots. These results are similar to those found by Day et al. [67] who found an increase in the abundance of the springtail Cryptopygus antarcticus under summer warming and water additions in the Antarctic Peninsula. However, mixed results have been common from warming experiments examining microarthropod abundance [68]. Bokhorst et al. [43] found little effect after 10 years of passive warming (OTCs) on invertebrate communities on Signy Island, while Convey et al. [69] showed significant declines in Collembola with warming as a result of desiccation. Ultimately, temperature-moisture regimes and microsite variation may have the most distinct influences on microarthropod communities in maritime Antarctica [67,69], and terrestrial site variation may be an important factor as invertebrates in wet sites may be differentially affected by warming experiments than those on drier or rocky sites in general.
While positive effects of warming were observed on microarthropod communities, moss species identity had a significant effect on abundance of soil nematodes. Our results show P. alpinum hosted a higher abundance of nematodes than S. georgicouncinata, demonstrating strong speciesspecific effects exist for the capacity of Antarctic mosses to host invertebrates. Mouratov et al. [70] found that the abundance of Antarctic soil nematodes on KGI was negatively correlated with soil water content. In addition, Mouratov et al. [70] found nematode abundance varied with plant host species, showing higher abundances under D. antarctica when compared with S. georgicouncinata, similar to findings in our study that revealed higher abundances under P. alpinum than S. georgicouncinata.
royalsocietypublishing.org/journal/rsos R. Soc. open sci. 6: 190744 The differential effects of warming on microarthropods and soil nematode abundance may be related to changing abiotic conditions in OTC chambers compared with control plots. Barcikowski and Loro [71] found species-specific differences in seasonal moss water content, showing that water content was both higher and less variable in P. alpinum when compared with S. georgicouncinata, further detailing how moss species can provide differential effects on the surrounding terrestrial environment. The availability of water and species-specific variation in moss water content and canopy temperatures we found with warming may influence both moss and soil fungi and microbial diversity, resulting in differential effects on microarthropod and soil nematode communities associated with bryophytes. In this study, we found that bryophyte fungal biomass was lower under passive warming, but also varied by species, which was contrary to our initial hypotheses. This decrease in mossassociated fungal biomass may be related to the significant increase in microarthropods associated with OTC warming, as both the moss-associated Oribatida and Collembola are known to feed on algae, dead organic material and fungi [72]. Clearly, more work is needed to better understand the effects of warming on bryophyte communities of the Western Antarctic Peninsula, including moss species-specific responses, both above and below ground, which may impact and alter the structure of invertebrate and microorganism communities and the function of this rapidly changing terrestrial ecosystem.

Species-specific effects, higher trophic levels and climate change
Research on bryophytes at higher latitudes suggests that mosses strongly influence ecosystem functions, including regulating water storage [21,25,[73][74][75] and temperature [58,76]. In Antarctica, ecosystem function among bryophytes has primarily been assessed in terms of nutrient inputs [77], while little is known about how mosses may also influence the thermal and water regimes that impact ecosystem function. The work in northern latitudes indicate that bryophytes are species specific with respect to their impacts on ecosystem function, such as thermal properties [29,30], although little is known about the ecosystem effects of such differences.
Our results demonstrate that Antarctic bryophytes have species-specific effects in thermal absorption and soil water availability, which may directly affect the plant community and higher terrestrial trophic levels. We found moss species-specific responses to passive warming, showing differential canopy temperatures and shifts in invertebrate communities between the two dominant moss species, P. alpinum and S. georgicouncinata. Furthermore, we observed that warming increased sexual expression of P. alpinum, which was also linked to microarthropod abundance, illustrating the broad range of species-specific biological responses in terrestrial Antarctica bryophyte communities. While we found significant effects of warming and species in our experiments, we did not find significant interactions between passive warming and moss species. This suggests that at least initially during climate warming, the relative differences among species may remain constant. Thus, first and foremost, we must understand the functional effects of Antarctic plants on this changing ecosystem; however, our results indicate that there is a reason to be hopeful that the functional effects that we measure today will be predictive under future climate change scenarios. Many challenges remain to understanding the impacts of rapidly changing environmental conditions on terrestrial communities in Antarctica, and a major task is the still insufficient knowledge about terrestrial biodiversity and how this unique biodiversity (dominated by cryptogams, microorganisms and microfauna) links to larger ecosystem function in light of continued climate warming [11,18]. Microarthropods play a fundamental role in nutrient cycling within soil ecosystems, yet in the Antarctic terrestrial ecosystem these processes have been shown to be largely governed primarily by climatic parameters [68,78,79]. Future studies to quickly enhance our species-specific understanding of the responses of cryptogam communities to warming will be the key to develop an ecosystem scale understanding of terrestrial ecology and ecosystem function in a warming Antarctica.