Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations

Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation... J. Hydrol. Hydromech., 69, 2021, 4, 421–435 ©2021. This is an open access article distributed DOI: 10.2478/johh-2021-0021 under the Creative Commons Attribution ISSN 1338-4333 NonCommercial-NoDerivatives 4.0 License Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations 1, ∞ 2, ∞ 3 4 2 2, * Sonja M. Thielen , Corinna Gall , Martin Ebner , Martin Nebel , Thomas Scholten , Steffen Seitz Invertebrate Palaeontology and Palaeoclimatology, Department of Geosciences, University of Tübingen, Schnarrenbergstr. 94-96, 72076 Tübingen, Germany. Soil Science and Geomorphology, Department of Geosciences, University of Tübingen, Rümelinstr. 19-23, 72070 Tübingen, Germany. Biogeology, Department of Geosciences, University of Tübingen, Hölderlinstr. 12, 72074 Tübingen, Germany. Nees-Institute for Biodiversity of Plants, University of Bonn, Meckenheimer Allee 170, 53115 Bonn, Germany. Corresponding author. Tel.: +49 (0)7071-29-77523. E-mail: steffen.seitz@uni-tuebingen.de These authors contributed equally to this project and are considered co-first authors. Abstract: Mosses are often overlooked; however, they are important for soil-atmosphere interfaces with regard to water exchange. This study investigated the influence of moss structural traits on maximum water storage capacities (WSC ) max and evaporation rates, and species-specific effects on water absorption and evaporation patterns in moss layers, moss- soil-interfaces and soil substrates using biocrust wetness probes. Five moss species typical for Central European temperate forests were selected: field-collected Brachythecium rutabulum, Eurhynchium striatum, Oxyrrhynchium hians and Plagiomnium undulatum; and laboratory-cultivated Amblystegium serpens and Oxyrrhynchium hians. –1 –1 WSC ranged from 14.10 g g for Amblystegium serpens (Lab) to 7.31 g g for Plagiomnium undulatum when im- max –1 –1 mersed in water, and 11.04 g g for Oxyrrhynchium hians (Lab) to 7.90 g g for Oxyrrhynchium hians when sprayed, due to different morphologies depending on the growing location. Structural traits such as high leaf frequencies and small leaf areas increased WSCmax. In terms of evaporation, leaf frequency displayed a positive correlation with evapora- tion, while leaf area index showed a negative correlation. Moisture alterations during watering and desiccation were largely controlled by species/substrate-specific patterns. Generally, moss cover prevented desiccation of soil surfaces and was not a barrier to infiltration. To understand water’s path from moss to soil, this study made a first contribution. Keywords: Biological soil crusts; Bryophytes; Ecohydrology; Moss structure; Moss hydrology; Rainfall interception. INTRODUCTION special water conducting cells (hydroids), the ectohydric movement of water is through spaces between adjacent shoots, Bryophytes occur in a wide range of ecosystems, from arctic leaves, leaves and stems, leaves and rhizoids and capillary and boreal enviroments to temperate and tropical forests, dry- systems such as leaf bases, revoluted leaf margins, grooves or lands, and even deserts (Hedenäs, 2007; Lindo and Gonzalez, networks of capillary channels determined by papillae 2010; Medina et al., 2011). They often form community assem- (Giordano et al., 1993; Glime, 2017). According to Schofield blages with other organisms such as lichens, fungi, algae, cya- (1981), capillary spaces are influenced by numerous structural nobacteria and bacteria, which form what are termed biological parameters such as leaf shape, leaf arrangement, leaf orienta- soil crusts (biocrusts) (Belnap et al., 2016). With approximately tion, detailed leaf anatomy (e.g. surface ornamentation), branch 20000 species, they are the second biggest group of land plants, arrangement, nature of cortical cells, and presence of rhizoids comprising mosses, liverworts and hornworts (Frey et al., 2009; or paraphyllia. Nevertheless, there is still limited data on moss Söderström et al., 2016). Moss layers fulfill crucial functional structural traits and water relations (Elumeeva et al., 2011). roles in a variety of ecosystems regarding water and nutrient Overall, mosses achieve maximum water storage capacities of fluxes (Bond-Lamberty et al., 2011; Cornelissen et al., 2007; 108% to 2070% of their dry weight (Proctor et al., 1998), with Gundule et al., 2011) as well as soil physical properties some Sphagnum species even reaching over 5000% of dry (Soudzilovskaia et al., 2013). In contrast to vascular plants, weight (Wang and Bader, 2018). mosses do not actively regulate their water content, but are Many mosses are capable of drying out without dying, poikilohydric, meaning their internal water content is in equi- which means they can endure losing all free intracellular water librium with ambient humidity (Green and Lange, 1994). For and recover their ordinary functions afterwards, such as photo- mosses, water is primarily available via rain, dew and fog synthesizing and growing when water is available (Proctor et (Glime, 2017) and moss moisture is influenced by many fac- al., 2007). Due to their high surface to volume ratios, rapid tors, depending on the habitat as well as the species itself in drying is generally facilitated (Proctor et al., 2007). Typically, moss cells are either completely turgid or desiccated, with regard to structure and life form (Dilks and Proctor, 1979; Oishi, 2018; Proctor, 1982; Proctor, 2000; Proctor and Tuba, relatively short transitions in between (Proctor et al., 2007). 2002), i.e. the form of individual moss shoots growing together, Factors influencing this water loss by evaporation are micro- which is considered an ecologically functional unit (Bates, climatic conditions (Proctor, 1990), life form characteristics 1998; Mägdefrau, 1982). (Elumeeva et al., 2011; Mägdefrau and Wutz, 1951; Nakatsubo, Water absorption occurs mainly via the external capillaries 1994; Zotz et al., 1997) and canopy structural properties such as (ectohydric), but in some species also via internal (endohydric) surface roughness, shoot density and cushion height (Goetz and movement. While the latter is achieved cell by cell or through Price, 2015; Rice and Schneider, 2004; Rice et al., 2001, 2018). Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz As an example of cushion life forms, Zotz et al. (2000) and METHODS Rice and Schneider (2004) found that evaporation rates de- Moss and soil characteristics crease with moss cushion size. For water balance of forest ecosystems, an intact forest floor Five moss species native to Southwest Germany (Nebel et cover such as leaf litter covers or moss layers play a crucial role al., 2001) differing in origin, classification and growth form (Acharya et al., 2017; Gerrits and Savenije, 2011; Mägdefrau were chosen for the study (Table 1). Oxyrrhynchium hians and Wutz, 1951; Sayer, 2006). In mid- and high-latitude conif- (Hedw.) Loeske, Eurhynchium striatum (Hedw.) Schimp., erous forests, moss layers often form at ground level (Elbert et Plagiomnium undulatum (Hedw.) T.J.Kop. and Brachythecium al., 2012). As forest ecosystems have suffered from drought in rutabulum (Hedw.) Schimp. were collected in the field at dif- recent years (Senf et al., 2020) and mosses are also increasingly ferent sites within the Ammer and Neckar valley. Cultures of threatened by global warming (He et al., 2016), it is particularly Amblystegium serpens (Hedw.) Schimp. and Oxyrrhynchium important to investigate their hydrological effects in these envi- hians were grown in a hydraulic fluid in an in vitro environ- ronments. Previous research by Price et al. (1997) in Canadian ment by Hummel InVitro GmbH in Stuttgart, Germany. The boreal forests showed that moss layers could retain 16.8 mm of latter was selected a second time to study intraspecific differ- water, which was approximately 21% of the precipitation input. ences between field and cultivated mosses. With regard to the Furthermore, Carleton and Dunham (2003) found that mosses position of the sporophytes, all selected mosses were pleuro- in a boreal forest could not be fully hydrated by capillary water carpous (side-fruited), except P. undulatum, which was acro- movement from the forest floor or dewfall, but rather from carpous (top-fruited). vapour from the forest floor condensing on the moss surface. Soil substrates were chosen according to common growing Liu and She (2020) investigated a linear decrease of soil evapo- conditions of selected moss species and sampled from four ration with increasing moss biomass, using moss that was pre- different sites in the Schönbuch Nature Park in Southwest viously cultivated in the laboratory. Overall, the forest floor Germany. Sampling sites were located in the geological series water balance is influenced by the amounts of throughfall rain, of the Lower Jurassic, with shale clay, interstratified by beds of the processes in the moss carpet, and the processes at the moss- pyrite and fine grained sandstone, as well as in the Upper Trias- soil interface (Price et al., 1997). However, little is known sic, where claystone with fine lime nodules and fine to coarse about how much water mosses release into the atmosphere and grained sandstone is present (Einsele and Agster, 1986). The how much is transported from the soil to the moss and vice substrates varied with regard to parent material, soil texture, versa (Glime, 2017; Voortman et al., 2014). In particular, the and pH as well as the C/N ratio (Table 2). They were sampled influence of different moss species on water movement through from the topsoil to a depth of 10 cm and sieved by 6.3 mm. Be- moss layers into the soil has been largely disregarded in this low, we distinguish the substrates according to their geological context, but has in turn shown great effects on e.g. erosion formation: Angulatensandstein (AS), Psilonotenton (PT), Löwen- control (Seitz et al., 2017). stein (LS) and Trossingen (TS) (Einsele and Agster, 1986). With this study, we aim to shorten this knowledge gap in an interdisciplinary approach (cf. Liu and She (2020)). To do so, Greenhouse experiment we examined water absorption and evaporation patterns in moss-covered soil substrates typical for a Central European With a greenhouse experiment, we investigated water ab- temperate forest during and after watering. We hypothesize that: sorption patterns in moss covers and corresponding soil sub- 1. Maximum water storage capacities (WSC ) of mosses strates during and after watering. To do this, we filled the sub- max are species-specific and positively affected by their surface area. strates into infiltration boxes (40 cm × 30 cm × 15 cm) up to a 2. Differences in the temporal dynamics of water content height of 6.5 cm. Infiltration boxes are stainless steel containers during watering and subsequent desiccation depend largely on with a triangular surface runoff gutter and an outlet on the the combination of moss species and the underlying soil sub- bottom to capture percolated water. In December 2019, moss strates. species were placed onto substrate-filled infiltration boxes, To test our hypotheses, we set up a greenhouse experiment leading to 6 treatments with 2 replicates each: P. undulatum with five moss species and four soil substrates, whereby (Field) + PT, O. hians (Field) + AS, O. hians (Lab) + AS, B. artificially cultivated mosses of the same species were also rutabulum (Field) + LS, A. serpens (Lab) + LS, E. striatum + included. We used biocrust wetness probes (Weber et al., 2016) TS; yielding a total number of 12 boxes. Infiltration boxes were for high-resolution monitoring of water content in moss layers, subsequently stored in a shady place outdoors for adaptation, on the soil surface, and in a soil depth of 3 cm. Furthermore, we until we began the greenhouse experiment in July 2020. investigated the selected mosses in terms of their structural traits and their maximum water storage capacities. Table 1. Characteristics of studied moss samples. Amblystegium Brachythecium Eurhynchium Oxyrrhynchium Oxyrrhynchium Plagiomnium serpens rutabulum striatum hians hians undulatum Family Amblystegiaceae Brachytheciaceae Brachytheciaceae Brachytheciaceae Brachytheciaceae Mniaceae Origin Lab Field Field Field Lab Field Site – ruderalized fertile pinewood dry hedge – flood plain characteristics meadow understore Growth form pleurocarpous pleurocarpous pleurocarpous pleurocarpous pleurocarpous acrocarpous Sample site – Tübingen Tübingen Reusten – Pliezhausen coordinates 48.544917 N 48.546194 N 48.541665 N 48.566723 N 9.043309 E 9.036407 E 8.914316 E 9.216494 E 422 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations Table 2. Characteristics of studied soil substrates. AS PT LS TS Series Lower Jurassic Lower Jurassic Upper Triassic Upper Triassic Formation Angulatensandstein- Psilonotenton-Formation Löwenstein-Formation Trossingen-Formation Formation (AS) (PT) (LS) (TS) Parent material sandstone shale clay sandstone claystone Soil texture silt loam silty clay loam clay loam silty clay loam • sand: 7.00 % • sand: 6.88 % • sand: 25.02 % • sand: 10.78 % • silt: 67.58 % • silt: 56.28 % • silt: 42.43 % • silt: 50.83 % • clay: 25.68 % • clay: 36.93 % • clay: 32.60 • clay: 38.10 % C/N 17.54 17.36 23.12 20.05 pH 5.8 7.0 7.0 5.6 Sample site Tübingen Tübingen Tübingen Tübingen coordinates 48.553054 N 48.557425 N 48.557527 N 48.556036 N 9.119053 E 9.114462 E 9.088098 E 9.089313 To measure water content (WC), we installed three biocrust used a gravimetrical approach with a heavy-duty precision bal- wetness probes (BWP; UP GmbH, Cottbus, Germany) per ance (KERN FCB 30K1, Kern & Sohn GmbH, Balingen, Ger- infiltration box in different positions: in 3 cm soil depth, in the many), and second, we used a Thetaprobe ML2 in combination uppermost 5 mm of the soil surface and in the moss layer (Fig. with a HH2 Moisture Meter (Delta-T Devices, Cambridge, UK). 1). BWPs were specifically developed to quantify WC of soil To consider evaporation effects during the period of desicca- surfaces as well as biocrusts by deriving WC from electrical tion, we calculated the evaporation rate of this time span for all conductivity measurements; they provided reliable data in samples using the formula several experiments under different field conditions (Gypser et WC – WC 0 x al., 2017; Löbs et al., 2020; Tucker et al., 2017; Weber et al., E = , 2016). Samples were irrigated for one hour with a sprayer t – t x 0 –1 (Comfort Sensitive Plant, Gardena, Ulm, Germany) with 6 L h where WC0 is the maximum gravimetric WC in the examined of water, split into 500 mL every 5 min, corresponding to a time period, WCx is the gravimetric WC at time point x, and tx precipitation amount of 122 mm (extremely heavy rainfall and t0 are the respective time points (Robinson et al., 2000). event). All BWPs were installed underneath the centre of the sprayer, whereby we ensured that the BWP in the moss layer Laboratory BWP calibration was completely encased by moss shoots. During this watering and subsequent desiccation process in the greenhouse, the elec- To calibrate the BWP to gravimetric WC, we monitored trical conductivity of the samples was logged every 10 seconds weight loss and electrical conductivity (EC) simultaneously for for 72 hours with the BWPs connected to a GP2 Data Logger all samples under laboratory conditions for at least 65 hours (Delta-T Devices, Cambridge, UK). Additionally, air tempera- (average air temperature: 19.1 °C, sd = 1.2 °C; average relative ture and relative humidity (RH) in the greenhouse were moni- humidity (RH): 45.8%, sd = 5.9%). Samples were water satu- tored (Tinytag Plus 2 – TGP-4500, Gemini Data Loggers, rated using the immersion technique described below (in the Chichester, UK) for the same time slots. Soil WC was deter- following section). Afterwards, they were placed on a balance mined before and after watering as well as after 71 hours of (Kern EW 620-3NM, Kern & Sohn GmbH, Balingen, Germa- desiccation applying two methods: first, we substrate inside ny) and three BWPs were installed in each sample. Two sam- ples were measured in parallel, using two precision balances of the same type. BWP and weight data were recorded at an inter- val of 10 seconds, while temperature and relative humidity were logged in 5 min intervals with Tinytag Plus 2 (see above). During monitoring of weight loss, the scales generated indi- vidual error values, which required a filtering of data. Since the scales only measured stable values, we had irregular time inter- vals in the recording of weight losses. To be able to combine weight and BWP as well as RH and temperature values, we performed a linear fashion interpolation with both weight val- ues and climate measurements. As EC is affected by temperature, we conducted a tempera- ture correction and derived the WC for a specific value of the BWP as described in Weber et al. (2016). According to Slatyer (1967), the formula (WW–DW) WC = DW Fig. 1. Overview of the greenhouse experiment setup. a) Biocrust –1 was used, where WC is the gravimetric WC (g g ), WW is the wetness probe (BWP) in 3 cm soil depth, b) BWP in soil surface wet weight (g) and DW is the dry weight (g) of the soil or moss and moss cover, c) experimental setup with moss-covered soil the sample. infiltration box and sprayer installed at uniform height. Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz The last step of calibration included curve fitting, where we then drained them for 2 min, and then weighed them. We used the mean of the three BWP values and the calculated WC. decided to use these two approaches, as we observed that some We found linear relationships which can be characterized as mosses were still dry on the bottom after a rainfall event, which WC = a · EC + b. For non-linear relationships we used non- was also described in Glime (2017). Therefore, we expected linear least-square regressions expressed by the equation different mechanisms of water absorption in the two WC = exp(a · ECt) · b · ECt + c, as recommended in Weber et al. techniques, with the spray technique probably being more (2016). Furthermore, some relationships could be better de- similar to the greenhouse watering process. The soil samples scribed with the equation WC = exp(a + b · ECt). While the were placed into a tub of water until the surface was wet and moss samples could be dried from saturation to desiccation, soil afterwards we measured the wet weight. To ensure that the soil samples did not dry out completely during the laboratory cali- substrate remained in the core cutter during rewetting, we bration. Therefore, an extrapolation of data for the calibration attached a thin water permeable fleece to the bottom of the core BWP values was necessary for the soil samples. An overview cutter (Blume et al., 2011). of all calibration curves is shown in Table S1 in the supporting information. Moss structural trait measurements Maximum water storage capacity To determine the surface areas of the studied moss species, we measured the following structural traits: leaf area, leaf fre- For a detailed characterisation of moss species and adjunct quency, shoot length, length of a single component (sum of soil substrates with regard to their maximum water storage shoot length and length of attached branches), shoot density capacity (WSCmax), further laboratory experiments were con- (Table 3). We determined the surface areas of the studied spe- ducted with samples from the infiltration boxes. Therefore, we cies using the following formula, which we adapted for our detached the mosses from the soil, dried them at 30 °C in a experiment following Simon (1987), Niinemets and Tobias dehydrator (Dörrex 0075.70, Stöckli, Netstal, Switzerland) and (2014) and Niinemets and Tobias (2019): weighed the dry samples (Mettler Toledo MS603S, Mettler leaf Toledo, Columbus, USA). Soil samples were taken with A = L N bryo shoot 1 cm shoot 100 cm metal core cutters from every infiltration box, dried at 105 °C in a compartment drier and weighed in dry state. After- where A is moss surface area, L is the average length of a bryo wards both moss species and soil substrates were saturated, shoot with its attached branches, N is mean number of shoot using two different methods for the mosses: spray and immer- measured shoots, and A is mean leaf area. Leaf area index leaf sion technique. For the spray technique, we moistened the (LAI) was then calculated with the formula mosses that had been placed in a petri plate with a spray bottle from above until samples could no longer absorb water. The bryo LAI = excess water was removed with a pipette and volume was de- sample area termined with a 25 mL measuring cylinder. By weighing the spray bottle before and after spraying we estimated the amount In the first step, three circular samples with a diameter of 5.5 of water added to the mosses (in average 3.45 mm). The wet cm (sample area) were taken from each species. Moss samples mosses were weighed again with the same balance. In contrast, were then dissembled into single moss shoots. Due to the very with the immersion technique we moistened the mosses by dense structure and consequent long time duration, only half of submerging them in water for 5 min between two soil sieves the circular area of A. serpens was considered. Next, detached with 52 µm mesh size on the bottom and 250 µm on the top, shoots were scanned using a high definition flatbed scanner Table 3. Species-specific average values (± standard error of the mean) of leaf area, leaf frequency, leaf area per shoot length, shoot length, length of a single component (sum of shoot length and length of attached branches), shoot density (shoot number per ground area), total surface area, leaf area index (LAI), moss cushion height, volume and density for the studied moss species. Species Leaf Leaf fre- Leaf area Shoot Length single Shoot Total LAI Cushion Cushion Cushion area quency per shoot length component density surface height volume density 2 –1 –2 2 3 –3 (mm ) (cm ) length (cm) (cm) (n cm ) area (cm ) (cm) (cm ) (g cm ) 2 –1 (cm cm ) Amblystegium 0.104 ± 81.778 ± 0.085 ± 1.168 ± 1.764 ± 97.005 ± 346.204 14.572 1.322 ± 107.058 ± 0.026 ± serpens (Lab) 0.002 3.929 0.006 0.024 0.224 11.786 0.091 10.623 0.002 Brachythecium 1.151 ± 39.333 ± 0.452 ± 3.791 ± 8.470 ± 3.031 ± 297.076 12.504 1.536 ± 139.856 ± 0.018 ± rutabulum 0.035 4.93 0.064 0.166 0.286 0.402 0.116 19.366 0.001 Eurhynchium 0.629 ± 91.333 ± 0.574 ± 2.018 ± 7.756 ± 2.511 ± 265.672 11.182 2.119 ± 182.071 ± 0.016 ± striatum 0.013 9.541 0.06 0.129 0.656 0.496 0.092 18.683 0.002 Oxyrrhynchium 0.307 ± 69.889 ± 0.187 ± 2.524 ± 8.124 ± 4.714 ± 169.907 7.151 1.65 ± 132.174 ± 0.015 ± hians 0.006 3.545 0.008 0.129 0.702 0.712 0.13 15.278 0.002 Oxyrrhynchium 0.393 ± 55.556 ± 0.219 ± 2.180 ± 6.198 ± 10.368 ± 333.764 14.048 1.353 ± 114.336 ± 0.022 ± hians (Lab) 0.008 2.911 0.014 0.092 1.480 2.509 0.136 18.998 0.003 Plagiomnium 4.737 ± 20.111 ± 0.953 ± 3.004 ± 4.960 ± 3.087 ± 346.517 14.585 1.394 ± 100.778 ± 0.018 ± undulatum 0.129 2.6 0.121 0.129 0.571 0.827 0.08 6.649 0.001 424 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations (Epson Perfection V700 Photo, Suwa, Japan) and shoot Moss structural traits numbers of all samples were counted to determine the shoot number per unit sample area. Afterwards, if sample size A wide range of structural trait characteristics for the moss enabled it, 50 shoots were randomly chosen for length species used in this study were determined to explain moss measurements, using ImageJ versions 1.53e and Fiji 2.1.0 water relations (Table 3). The average individual leaf area of (Schindelin et al., 2012; Schneider et al., 2012). Next to shoot the studied species ranged almost fivefold from 0.104 mm² in length, we also determined the length of branches that were A. serpens (Lab) to 4.737 mm² in P. undulatum. Accordingly, attached to the measured shoots. Then, from each sample three average leaf area per shoot length varied elevenfold between 2 –1 shoots were randomly selected and all leaves were carefully 0.085 cm cm in A. serpens (Lab) to 0.953 in P. undulatum. removed along one centimeter of the shoot. The removed leaves Leaf frequency was the smallest in P. undulatum at 20.111 and were put on slides and were either scanned with the flatbed ranged up to 91.333 in E. striatum. We found the longest shoots scanner or a digital microscope (Keyence in B. rutabulum (3.79 cm on average) and the shortest shoots in VHX-7000 with dual zoom lens VH-ZST, Keyence, Osaka, A. serpens (Lab) (1.16 cm on average). After adding the length Japan). Leaf area was subsequently measured with ImageJ as of attached branches to the respective shoot length, B. rutabu- well. lum still had the longest shoots with 8.47 cm, and A. serpens Additionally, we determined the volume of the moss (Lab) had the shortest shoots with 1.764 cm. However, A. ser- cushions for all moss samples used in the WSCmax experiment. pens (Lab) had the highest shoot density (97 shoots per cm ), Therefore, we photographed all moss samples using a Nikon whereas B. rutabulum, E. striatum, O. hians and P. undulatum D5100 (Chiyoda, Japan), equipped with an AF-S DX Micro had much lower densities between 2.5 to 4.714 shoots per cm . NIKKOR 40mm f/2.8G lens to identify the individual sample Interestingly, shoot density of O. hians (Lab) was twice as high area with ImageJ. The height of the moss cushions was as for O. hians collected in nature, which might be due to miss- measured at four sites with a calliper and mean values were ing competition with other species in a laboratory setting, as calculated for every cushion. The moss cushion density was well as different light and water regimes, since moss structure derived from the quotient of dry weight and cushion volume. is highly affected by water and light availability (Mägdefrau, 1982). This raises the question of whether field-collected A. Data analysis serpens also has similarly high shoot densities as determined for A. serpens (Lab) in this study. While A. serpens (Lab) grew All analyses were conducted with R software versions 3.6.3 in dense and more voluminous lawns, A. serpens occurs more and 4.0.2 (R Core Team, 2021) on the level of individual sam- often intermingled with other species in nature. The nutrient- ples. To examine significant differences, we used one-way loving species prefers semi-shady, rather moist sites that are ANOVAs in combination with post-hoc Tukey’s HSD tests also preferred by many other species that are often more vigor- when variables showed homogeneity of variances. In other ous and thus overgrow the delicate prostrate A. serpens (Nebel, cases, we performed post-hoc Games-Howell or Wilcoxon 2001). The dense, extensive tall lawns of A. serpens (Lab) signed-rank tests. Previously, homoescedasticity was verified therefore contradict the species’ occurrence in nature and its with the Levene’s test. To test for differences of the means interspersed growth with other mosses, that can be attributed to between two samples we used Welch’s t-test. Significance was the low competitiveness of A. serpens. assessed at p < 0.05 in all cases. Compared to the other five studied species, O. hians had a Furthermore, we performed pairwise Pearson as well as low LAI of 7.151. B. rutabulum and E. striatum were similar in Spearman’s Rank correlation analyses to screen for relation- their LAI of 12.504 and 11.182, respectively, and highest LAI ships between WSC as well as evaporation rates of the stud- values were determined for A. serpens (Lab) (14.572), O. hians max ied samples and parameters of sample characteristics. In ad- (Lab) (14.048) and P. undulatum (14.585). Interestingly, P. vance of all analyses, we used the Shapiro-Wilk Test to exam- undulatum and the two lab-grown mosses are very different in ine the samples for normal distribution. Additionally, general- terms of leaf area, leaf frequency and shoot density, but all have ized additive models (GAM) with restricted maximum likeli- similar LAI values. Considering the moss cushion density, A. hood and smoothing parameters selected by an unbiased risk serpens (Lab) was significantly denser than E. striatum (p < estimator (UBRE) criterion were performed to assess the effect 0.001), O. hians (p < 0.001) and P. undulatum (p < 0.01). Fur- of soil substrate or moss species characteristics on WSCmax. thermore, we found significant differences in regard to moss Firstly, we fitted moss WSCmax from the spray and immersion cushion density between O. hians (Lab) and E. striatum (p < techniques against mean shoot number, mean leaf surface area, 0.01), O. hians (Lab) and O. hians (p < 0.05), B. rutabulum and LAI, moss cushion height as well as moss cushion density. E. striatum (p < 0.05) and E. striatum and P. undulatum (p < Secondly, WSC of soil substrates were fitted against soil 0.05). max bulk density, sand, silt and clay contents as well as total carbon and nitrogen content. Maximum water storage capacity RESULTS AND DISCUSSION Mean values of WSC from the immersion technique (rep- max –1 resenting complete soaking) varied between 14.10 g g for A. –1 In order to discuss and answer the hypotheses presented, we serpens (Lab) and 7.31 g g for P. undulatum, with the differ- first analyzed the differences in structural traits of the studied ence being highly significant (p < 0.001) (Fig. 2 and Table S2 moss species and investigated their relationship with WSCmax. in the supporting information). Further significant differences –1 As we assumed that the temporal progression of WC in the were found between E. striatum (11.22 g g ) and P. undulatum –1 greenhouse experiment could be explained by the structural as well as between B. rutabulum (11.80 g g ) and P. undulatum traits of moss species, we further examined whether our sam- (p < 0.05). Thus, with regard to the WSCmax, there were strong ples showed similar patterns of properties in the different ex- differences between the mosses with different growth forms, periments. but none within the group of pleurocarpous mosses. The fact Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz –1 Fig. 2. Maximum water storage capacity (g g ) of treatments (moss species + soil substrate). For moss species both spray and immersion technique are illustrated. Crosses represent mean values and lines within boxplots median values. The bottom and top of the box represent the first and third quartiles, and whiskers extend up to 1.5 times the interquartile range (IQR) of the data. Outliers are defined as more than 1.5 times the IQR and are displayed as points. that P. undulatum absorbed comparatively less water could be importance to the WSC . Such parameters are assumed to be, max explained by its endohydric water transport, and many acrocar- for example, the capillary spaces of mosses, which are very pous mosses are endohydric (Richardson, 1981). Since the difficult to quantify and are diverse and often complex (Proctor, surface of endohydric mosses comprises a water-resistant cuti- 1982). According to Proctor (1982), capillary conducting sys- cle with often waxy layers (Buch, 1945; Proctor, 1979a; tems such as spaces between overlapping leaves, between 1979b), water absorption through their leaves is inhibited shoots, in sheathing leaf bases or amongst rhizoid tomentum (Glime, 2017). However, as we only measured one acrocarpous and paraphyllia can be 10–100 µm large. In addition, interspac- moss, this finding requires further investigation. es of a few µm can be found in interstices between papillae as Although the most significant difference in WSCmax was well as in furrows between plicae and ridges on leaves and shown between the visibly densest and loosest growing moss stems (Proctor, 1982). In this context, the 3D structure of the species, this relationship could not be substantiated by the mosses, e.g. the branching of the shoots, the shape of the leaves surveyed traits for surface area and cushion characteristics. and the position of the leaves in relation to the stems, potential- WSC was not affected by total surface area or LAI. Further- ly plays an important role for capillarity of bryophytes max more, neither height of the moss cushions, nor volume or densi- (Giordano et al., 1993; Schofield, 1981). ty correlated individually with WSC . The combination of In contrast to the immersion technique, the range of mean max leaf area and leaf frequency seemed to have a higher influence values of WSC for the spray technique, which was intended max on WSC : with a small leaf area (Spearman’s correlation to simulate moistening of mosses by a rainfall event, was con- max rho = –0.30, p < 0.05) and high leaf frequency (Spearman’s siderably smaller (Fig. 2). Here, we found a variation of 11.04 –1 –1 correlation rho = 0.32, p < 0.05), the WSCmax increased. Shoot g g for O. hians (Lab) to 7.90 g g for O. hians from the field. density might be another influencing factor, but due to small However, we could not find any significant differences between sample size further studies are recommended. In this context, species or significant correlations between the WSCmax and the Voortman et al. (2014) also discussed that capillary spaces ascertained individual moss structure parameters, and the ad- between moss leaves and branches might be more relevant for junct GAM could explain 46.5% of the deviations. The greatest water retention than those between moss shoots. For Sphagnum influence was due to moss height (p < 0.01), with LAI having a species, Bengtsson et al. (2020) also found a high influence of smaller effect (p < 0.05). Interestingly, the greatest difference leaf traits on water retention. in WSC was discovered within the same species, O. hians. max Calculated in a GAM explaining 54.1% of the deviance, Although they belong to the same species, O. hians collected in moss cushion density highly influenced WSC (p < 0.001), the field and O. hians grown in the laboratory displayed strong max while the effects of mean leaf area (p < 0.01) and mean shoot differences in structure. While O. hians grows as loose lawn density (p < 0.05) were smaller, but also significant. Therefore, in the field, the laboratory variety forms very dense moss we assume that additional parameters must be also of great cushions, which is also reflected in the higher shoot density 426 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations (O. hians: 4.714 shoots per cm and O. hians (Lab): 10.368 with bulk density (Pearson’s correlation r = –0.70, t = –5.94, p shoots per cm ), and the larger total surface area (O. hians: < 0.001) and C/N ratio (Spearman’s correlation rho = –0.62, p 2 2 169.907 cm and O. hians (Lab): 333.764 cm ). This finding < 0.001). Additionally, we tested for a joint impact on soil indicates that the WSC of mosses is dependent on life form. WSC using a GAM with soil bulk density, sand, silt and clay max max In a further chain of thought, this also implies that single spe- contents as well as total carbon and nitrogen content as fixed cies can obtain more advantageous properties through laborato- effects and were able to explain 84.7% of the deviance with this ry cultivation, e.g. for erosion control. model. The results also showed a high relevance of bulk soil Overall, we suppose that for both the immersion technique density as well as total carbon content (p < 0.001), which is and the spray technique, the capillary spaces between moss consistent with the results of the individually tested correlations shoots as well as between leaves and shoots are more important and an influence of the clay content (p < 0.01). These relation- for WSC than surface parameters such as LAI or total sur- ships are also reported in other studies (Gong et al., 2003; max face area. Finally, it can be concluded that a further develop- Franzluebbers, 2002; Novák and Hlaváčiková, 2019; Rawls et ment and standardization of the spray technique is required to al., 2003). be able to gather more reliable data on this important moss characteristic. Greenhouse experiment Regarding the soil substrates, WSC values varied on av- Watering process max –1 –1 erage between 0.46 g g for PT and 0.36 g g for TS, which is 30 times less compared to the WSCmax of the mosses (Fig. 3). Focusing on the 60 minutes of watering, we observed clear Within soil substrates we found highly significant differences differences in WC of different moss species, regarding tem- between PT and TS, PT and LS as well as AS and TS poral progression as well as the level of WC achieved (Fig. 3). (p < 0.001) and a significant difference between AS and TS At the beginning of the watering, all moss species were desic- (p < 0.05). On one hand, these differences can be explained by cated, so that the WC initially increased until an equilibrium soil texture, as there is a negative relationship with sand content was reached. Moss species were classified in terms of WC in –1 equilibrium: (a) low WC (0–5 g g (Spearman’s correlation rho = –0.62, p < 0.001) and a positive ) for A. serpens (Lab) and P. –1 correlation with silt content (Spearman’s correlation rho = 0.52, undulatum, (b) medium WC (5–10 g g ) for B. rutabulum, O. –1 p < 0.001), while the clay content seemed to be of rather minor hians and O. hians (Lab), (c) high WC (10–15 g g ) for E. importance for WSC (Spearman’s correlation rho = –0.40, p striatum. This classification shows the possibility of distin- max < 0.01). On the other hand, we revealed a negative correlation guishing between moss species based on the BWP response. -1 Fig. 3. Temporal progression of water content values (g g ) of treatments during watering in the greenhouse experiment. Replicate measurements are labelled with A and B for every biocrust wetness probe (BWP) location (moss cover, soil surface, 3 cm soil depth). Plotted are half-minute values. Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz Surprisingly, A. serpens (Lab) and P. undulatum both reached a Compared to the mosses, the soil substrates showed a much low WC during irrigation, although they are quite different re- lower WC during the 60 minutes of watering, which is true for garding their structural traits. While A. serpens (Lab) forms very both the surface and 3 cm soil depth (Fig. 3). Overall, mosses –1 dense moss cushions (shoot density: 97.005 ± 11.786 shoots per adjusted their equilibrium in the range between 2.5–15.0 g g 2 –1 cm ), P. undulatum is more likely to grow single shoots (shoot of WC, while soil substrates varied between 0.15–0.35 g g . density: 3.087 ± 0.827 shoots per cm ). Although O. hians and O. The fact that mosses can absorb more water than soil substrates hians (Lab) were both assigned to medium WC, we recognized a could be attributed to a larger surface area of mosses. Addition- –1 distinct difference, with O. hians tending to weigh 10 g g and O. ally, capillary effects in mosses might contribute to higher –1 hians (Lab) tending to weigh 5 g g . Since O. hians (Lab) grows water absorption rates compared to soil substrates. considerably denser than O. hians with a shoot density twice as Since the soil surfaces were not completely dried out at the high (O. hians (Lab) = 10.368 ± 2.509 shoots per cm , O. hians = beginning of the experiment, they showed a relatively high 4.714 ± 0.712 shoots per cm ) and a higher cushion density (O. starting value of WC in comparison with the later reached hians (Lab) = 0.022 ± 0.003 shoots per cm , O. hians = 0.015 ± equilibrium. The temporal progression of WC on the soil sur- 0.002 shoots per cm ), we expected that O. hians (Lab) would face started with higher values at the beginning of watering and also absorb more water during watering. The fact that this ex- slightly decreased over time. Regarding infiltration into the soil pectation was not fulfilled could be attributed to O. hians hav- surface, it appeared that water had initially accumulated on the ing a comparatively high leaf frequency with small leaf area, surface, causing the high WC. which had already been highlighted as important factors for When considering WC at 3 cm soil depth, temporal progres- water absorption in previous chapters. sion of WC was almost steady, which was also due to the al- Furthermore, almost all moss species showed a certain varia- ready wet soil substrate at the beginning of the experiment. For tion in WC at equilibrium within replicate measurements, illus- two substrates (AS and PT) we observed an increase of WC trating a great heterogeneity within species. Overall, we noticed during the first 10 minutes of irrigation, indicating percolation that the variations between replicate measurements were small- of water through the substrate. Additionally, WC tended to be er for denser moss cushions than for looser ones, with P. undu- lower at 3 cm soil depth than on the soil surface during irriga- latum being an exception in this case. This could be attributed tion. Overall, with respect to the temporal progression of WC to the fact that denser mosses establish better contact with the values on soil surface and in 3 cm soil depth, we generally sensor without forming air spaces (Löbs et al., 2020). found substrate-specific coherences regarding the level of WC Some moss species demonstrated a more pronounced re- achieved. sponse to the watering pulses than others. This might also be Furthermore, we expected that the soil substrates show a related to denser moss cushions with less air-filled interstitial similar response due to WSC in the lab and in the green- max spaces (Löbs et al., 2020), as it was the case for A. serpens (Lab) house experiment. However, the WC after watering in the and O. hians (Lab), which both form the densest cushions. To greenhouse, which we expected to be the maximum WC examine moss intraspecific differences regarding water absorp- reached in the greenhouse (means of WC for all values between th th tion in detail, higher replication is necessary in future studies. 60 and 65 minute), were lower than the WSCmax measured in Because of the water volume applied to the samples in the the lab, which was true for every substrate both for surface as greenhouse, we speculated that the moss species would reach well as in 3 cm soil depth. For example, PT achieved a WSCmax –1 –1 their WSCmax within the watering time in the greenhouse, espe- of 0.46 g g and only showed a WC of 0.31 g g on the surface –1 cially when compared with the achieved WSCmax using the and 0.27 g g in the soil after one hour of watering in the spray technique. To go into more detail, we compared the WC greenhouse, which means a deviance of 32.61%. In compari- values directly after watering (means of WC for all values son, LS reached only 50% of the WSC under A. serpens max th th –1 –1 between 60 and 65 minute) with the WSC determined in (Lab) (WC after watering = 0.19 g g , WSC = 0.46 g g ) max max the lab. For most of the species the WC after watering was and 52% under B. rutabulum on the surface (WC after watering –1 considerably lower than the WSC , for both spray and immer- = 0.18 g g ), WC values in 3 cm soil depth were even lower max –1 sion technique. As an example, the maximum WC for A. ser- (WC after watering (A. serpens (Lab)) = 0.18 g g ; WC after –1 pens using the immersion technique was 5 times higher than the watering (B. rutabulum) = 0.17 g g ). Altogether, soil sub- –1 WC after watering (WSCmax (immersion) = 14.10 g g , SE = strates did not show the same patterns of water absorption in –1 1.28, WC after watering = 2.63 g g , SE = 0.02), while the the lab as in the greenhouse. spray technique showed almost a fourfold difference (WSCmax –1 (spray) = 10.10 g g , SE = 1.25). Additionally, we found an Desiccation process almost fivefold difference from the immersion technique, re- spective fourfold difference from the spray technique, and During the subsequent desiccation process of 71 hours, higher WSCmax compared to the WC after watering in P. undu- moisture in the moss layers generally decreased, while moisture –1 latum (WSC (immersion) = 7.31 g g , SE = 0.80, WSC at the soil substrate surface as well as in 3 cm soil substrate max max –1 –1 (spray) = 8.15 g g , SE = 0.32, WC after watering = 1.76 g g , depth remained at the same levels (Fig. 4). However, moss SE = 0.01). Based on these results, no clear patterns are species differed in maximum WC, evaporation rates and their discernible that would explain the different intraspecific mech- responses to climatic changes in the greenhouse. Sample repli- anisms of water absorption comparing greenhouse and labora- cates slightly differed from each other in regard to WC values, tory experiments. Above all, it was very surprising that espe- but generally showed comparable patterns. We observed the cially the denser mosses, most notably the lab-grown mosses, highest WC values directly after watering in E. striatum with a –1 did not absorb much water during the greenhouse experiment. mean WC of almost 15 g g , while mean WC of B. rutabulum, –1 In general, we can deduce that the mosses are not a barrier to O. hians and O. hians (Lab) ranged between 5–10 g g , and infiltration in case of high precipitation rates, as also reported in mean WC of A. serpens (Lab) and P. undulatum did not exceed –1 Li et al. (2016). A new observation of our study is that the 5 g g . The low WC of P. undulatum might be related to its mosses growing on the soil do not store much of the applied delicate and loose structure with a low leaf frequency and large water themselves, but pass it on to the soil. leaf areas, and leaves that stand off the shoot. Especially 428 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations –1 Fig. 4. Temporal progression of water content values (g g ) of treatments during 71 h of desiccation in the greenhouse experiment. Repli- cate measurements are labelled with A and B for every biocrust wetness probe (BWP) location (moss cover, soil surface, 3 cm soil depth). Mean temperature and mean relative humidity ± standard deviation: Amblystegium serpens (Lab) + Löwenstein-Formation (LS) 25.93 ± 6.13 °C, 42.67 ± 14.39%; Brachythecium rutabulum + Löwenstein-Formation (LS) 26.35 ± 5.38 °C, 49.45 ± 15.22%; Eurhynchium stria- tum + Trossingen-Formation (TS) 24.70 ± 5.78 °C, 46.31 ± 16.15%; Oxyrrhynchium hians + Angulatensandstein-Formation (AS) 20.30 ± 3.89 °C, 64.72 ± 18.45%; Oxyrrhynchium hians (Lab) + Angulatensandstein-Formation (AS) 23.10 ± 6.07 °C, 53.37 ± 18.38%; Plagiomni- um undulatum + Psilonotenton-Formation (PT) 20.96 ± 4.31 °C, 59.92 ± 15.35%. Plotted are hourly values. –1 compared to a branched structure with high leaf frequencies ably smaller (0.023 and 0.012 g h ). A group with slightly and densely attached leaves, few capillary spaces for water higher evaporation rates consisted of A. serpens (Lab) (0.056 –1 –1 storage are formed in P. undulatum (Mägdefrau and Wutz, and 0.03 g h ), B. rutabulum (0.046 and 0.055 g h ), O. hians –1 1951). Furthermore, leaf surfaces of mosses from the Mniaceae (Lab) (0.057 and 0.078 g h ) and O. hians (0.06 and 0.093 –1 family often have a water-resistant cuticle, reducing their g h ). We found a positive relationship between leaf frequency ability to absorb water via the leaves (Glime, 2017; Proctor, and evaporation rate (Spearman’s correlation rho = 0.832, P < 2000). Additionally, we observed that leaves and stems of P. 0.001). LAI, however, correlated negatively with evaporation undulatum were twisting and curling during the desiccation rate (Spearman’s correlation rho = –0.78, P < 0.001); this was process, which might result in altered measurement conditions congruent with our expectations of lower evaporation rates for for the sensor. Clipping the sensor to moss stems of such moss species with a high LAI, which, as a product of different species as P. undulatum, as proposed in Leo et al. (2019), structural traits, makes the formation of a multitude of capillary would be interesting to compare with BWP response in future spaces for water storage in different hierarchical levels (leaf, studies. Nevertheless, the BWP used in this study proved to be shoot, and colony) more likely, overall resulting in wetter moss successful in all moss species, as also confirmed in Löbs et al. cushions and lower evaporation rates, as also described in (2020). Elumeeva et al. (2011). A. serpens (Lab) had dried out after 30 hours, whereas the WC in moss species showed diel fluctuations, albeit to other species generally remained moist longer than 40 hours, different degrees. Desiccation periods clearly aligned with and did not desiccate completely during the measurement. A declining RH and rising temperatures in E. striatum, O. hians more stabilized, steady evaporation was observed in B. rutabu- (Lab), and to a smaller degree in P. undulatum, A. serpens lum, O. hians, O. hians (Lab) and P. undulatum. Evaporation (Lab), B. rutabulum and O. hians. Comparably high RH and rates calculated for the measurement period corresponded to low temperatures contributed to the quite stable WC of O. hians maximum WC: E. striatum with the highest maximum WC throughout the measurement and to the fact that the moss did after watering also had the highest evaporation rates (0.181– not dry completely. We observed slight reactions of WC –1 0.197 g h ). Evaporation rates for P. undulatum were consider- towards RH changes in all samples, confirming that mosses Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz reacted to increasing RH and could absorb water under CONCLUSIONS AND OUTLOOK conditions with high RH, as also described in Löbs et al. (2020). This study found that five moss species from Central Euro- Climatic conditions cannot explain intraspecific variation of pean temperate forests can exhibit different water absorption WC, since the replicates were measured in parallel at the same and evaporation patterns in response to rainfall. In some cases, time. A possible explanation could be that moss structure at the the target moss species also showed significant intraspecific sensor locations differed in regard to surface roughness, alter- variability in rainwater interception. With regard to our hypoth- ing boundary layer resistance and thus resulting in different eses, the following conclusions were drawn: evaporation velocities (Proctor, 1982). Further experiments in a 1. Contradictory to our hypothesis, total surface area did climate-controlled environment with closer control and manipu- not affect maximum water storage capacity (WSCmax). Results lation possibilities could determine if moss reactions are spe- further indicate that a combination of structural traits (high cies-specific. shoot density, high leaf frequency, and low leaf area) may The different soil substrates had slightly different mean WC increase WSC during immersion. Generalized additive max –1 –1 values in 3 cm depth: LS 0.16–0.18 g g , TS 0.24 g g , AS models (GAM) revealed that cushion density also can influence –1 –1 0.28 g g and PT 0.24 g g . In LS, a slight reaction to rising WSC . A combination of different structural traits tested in a max RH (due to night-day-shifts) was recognizable, and LS did not GAM showed that WSC determined using the spray max desiccate, despite high temperatures above 40 °C during the technique was affected by leaf area index (LAI) and moss measurement period. We assume that the moss cover prevented height. Overall, soil substrates absorbed around 30 times less desiccation of the substrate, but it remains unclear whether the water compared to mosses and an effect of bulk density, grain substrate receives water from the moss cushion itself or plainly size distribution and total carbon content on WSCmax was from RH. For low precipitation rates, prevention of soil evapo- found. ration from moss-dominated biocrusts was also reported in Li et 2. Both moss species and soil substrates showed al. (2016). species/substrate-specific patterns in regard to changes of WC at the soil surface fluctuated diurnally depending on RH moisture during watering as well as desiccation. Since soil as also described in Tucker et al. (2017), especially in AS and substrates did not desiccate despite high temperatures, yet water PT and less pronounced in LS. Moreover, we found that content at the surface responded to relative humidity changes, oscillations related to RH were visible at the soil surface but not we hypothesize that the moss cover prevented desiccation in 3 cm soil depth, which showed that fine pores at the surface without sealing the soil. Because the humidity-induced were capable of adsorbing water out of the air (Agam and fluctuations varied depending on the density of the moss cover, Berliner, 2006; Hillel, 1998). So even dense moss cushions we further hypothesize that mosses mitigate soil evaporation. were not completely sealing the soil surface and there was no Among moss species, differences were also observed between full barrier by bryophytes. However, since the RH-induced replicates, primarily related to the moistening until an fluctuations varied depending on the density of the moss cover, equilibrium in water content was reached, as well as in the i.e. the most pronounced reactions were found in the loosest process of desiccation. Similar WSCmax values (to immersion moss cover P. undulatum, we assume that mosses mitigate soil and spray) were not achieved in greenhouse experiments during evaporation. watering, indicating different mechanisms of water absorption Generally, WC at the soil surface was higher than in 3 cm for both soil substrates and moss species, which could not be depth during desiccation. This could be ascribed to the fact that explained by clear patterns. In general, we can deduce that the the soil surface had a finer texture due to clogging of the pores mosses growing on the soil may not store much of the applied as an influence of splash effects (Morgan, 2005), which allows water themselves, but pass it on to the soil. Leaf frequency for a higher WC (Dodd and Lauenroth, 1997). A further influ- correlated positively with evaporation rates, while LAI showed encing factor to explain this observation might be the initial soil a negative relationship with evaporation rates. WC. As we measured a high soil WC before watering, the Although not explicitly mentioned in our hypotheses, the matrix potential is reduced, resulting in a lower and less deep results underscore that some species can develop different infiltration (Novák and Hlaváčiková, 2019). morphologies due to different growing locations (field vs. la- Differences between WC values of surface and 3 cm depth boratory). This can lead to a heterogeneous expression of the depended on the substrate: for LS, the values were very similar, same traits and raises the question of whether beneficial traits but especially for PT, WC values at substrate surface were can be conferred to individual species by laboratory cultivation, higher than in 3 cm depth by a factor of 1.4 to 2.3. In AS, there e.g. for erosion control. Thus, the interplay of individual moss was either an influence by the moss cover, or by the climatic structure traits appears to be very complex, such that further conditions during the measurement: AS covered with O. hians detailed investigations especially on the 3D structure of indi- showed a smaller difference between surface and soil WC and vidual species are urgently needed. In this context, more infor- not very pronounced oscillations with RH. In contrast, AS mation on moss capillary spaces would help to achieve a higher covered with O. hians (Lab) displayed strong day-night level of accuracy regarding the mechanisms of water absorption oscillations and WC values during nights were up to 1.5-times in mosses. It should be noted that the methodology also needs higher in the surface than in 3 cm depth. Since RH remained further improvement and the exact determination of individual above 50% after 20 hours during the measurement of AS with species effects can be seen as non-trivial. O. hians, but dropped from 75% in the nights to 25% RH Considering that the methodology has proven to be sound, during the days in the measurement of AS with O. hians (Lab), the full significance of the current results in this study needs to we cannot exclude a strong influence of these fluctuations on be confirmed in a larger experimental setup. Further research is the different oscillation patterns in the AS measurements. To required to understand the details of how different moss species determine the effect of moss layers itself on soil substrate and soil substrates interact regarding water absorption and moistness and evaporation, an experiment with different moss evaporation. A multi-method approach to measure water con- species on similar substrates and control samples without moss tent in different layers is recommended, using biocrust wetness is necessary. probes as well as clip sensors for the moss cover as introduced 430 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations by Leo et al. (2019). This approach should be combined with Morphologie des Schönbuchs. In: Einsele, G. (Ed.): Das the use of a climate-regulated greenhouse and expanded to landschaftsökologische Forschungsprojekt Naturpark include control samples without moss cover and large number Schönbuch: Wasser- und Stoffhaushalt, Bio-, Geo- und of replicates in order to cover the existing complexity as well as Forstwirtschaftliche Studien in Südwestdeutschland. VCH possible. This complexity is also the major challenge in the Verlagsfesellschaft, Weinheim. investigation of "water’s path from moss to soil", to the under- Elbert, W., Weber, B., Burrows, S., Steinkamp, J., Büdel, B., standing of which this study has made a further contribution. Andreae, M.O., Pöschl, U., 2012. Contribution of cryptogamic covers to the global cycles of carbon and Acknowledgements. We thank Michael Sauer for his assistance nitrogen. Nature Geoscience, 5, 459–462. and expertise during moss sample collection, Lena Grabherr Elumeeva, T.G., Soudzilovskaia, N.A., During, H.J., and Sarah Fodor for their support in measuring moss structural Cornelissen, J.H., 2011. The importance of colony structure parameters, Daniel Schwindt for helpful comments and Mat- versus shoot morphology for the water balance of 22 thew Hughes for improving the language of the manuscript. We subarctic bryophyte species. Journal of Vegetation Science, are also grateful to the Plant Ecology group of the University of 22, 152–164. Tübingen for the space provided in their greenhouse. We thank Franzluebbers, A.J., 2002. Water infiltration and soil structure three anonymous reviewers for their constructive comments. related to organic matter and its stratification with depth. This study was funded by the German Research Foundation Soil and Tillage Research, 66, 197–205. (DFG SE 2767/2-1 “MesiCrust”) and the Federal Ministry for Frey, W., Stech, M., Fischer, E., 2009. Syllabus of Plant Economic Affairs and Energy (AiF 19808 N). Families - Part 3: Bryophytes and Seedless Vascular Plants. Borntraeger, Berlin, Stuttgart. REFERENCES Gerrits, A.M.J., Savenije, H.H.G., 2011. Forest floor interception. In: Levia, D.F., Carlyle-Moses, D., Tanaka, T. Acharya, B.S., Stebler, E., Zou, C.B., 2017. Monitoring litter (Eds.): Forest Hydrology and Biogeochemistry: Synthesis of interception of rainfall using leaf wetness sensor under Past Research and Future Directions. Springer Netherlands, controlled and field conditions. Hydrological Processes, 31, Dordrecht, pp. 445–454. 240–249. Giordano, S., Colacino, C., Spagnuolo, V., Basile, A., Esposito, Agam, N., Berliner, P.R., 2006. Dew formation and water A., Castaldo-Cobianchi, R., 1993. Morphological adaptation vapor adsorption in semi-arid environments – A review. to water uptake and transport in the poikilohydric moss Journal of Arid Environments, 65, 572–590. Tortula ruralis. Giornale Botanico Italiano, 127, 1123–1132. Bates, J.W., 1998. Is 'life-form' a useful concept in bryophyte Glime, J.M., 2017. Volume 1: Physiological Ecology. ecology? Oikos, 82, 223–237. Bryophyte Ecology. Belnap, J., Weber, B., Büdel, B., 2016. Biological soil crusts as Goetz, J.D., Price, J.S., 2015. Role of morphological structure an organizing principle in drylands. In: Weber, B., Büdel, and layering of Sphagnum and Tomenthypnum mosses on B., Belnap, J. (Eds.): Biological Soil Crusts: An Organizing moss productivity and evaporation rates. Canadian Journal Principle in Drylands. Springer International Publishing, of Soil Science, 95, 109–124. Cham, pp. 3–13. Gong, Y., Cao, Q., Sun, Z., 2003. The effects of soil bulk Bengtsson, F., Granath, G., Cronberg, N., Rydin, H., 2020. density, clay content and temperature on soil water content Mechanisms behind species-specific water economy measurement using time-domain reflectometry. responses to water level drawdown in peat mosses. Annals Hydrological Processes, 17, 3601–3614. of Botany, 126, 219–230. Green, T.G.A., Lange, O.L., 1994. Photosynthesis in Blume, H.-P., Stahr, K., Leinweber, P., 2011. Bodenkundliches poikilohydric plants: a comparison of lichens and Praktikum: eine Einführung in pedologisches Arbeiten für bryophytes. In: Schulze, E.-D., Caldwell, M.M. (Eds.): Ökologen, insbesondere Land- und Forstwirte, und für Ecophysiology of Photosynthesis. Springer, New York, pp. Geowissenschaftler. 3. neubearbeitete Auflage edn. 319–341. Spektrum Akademischer Verlag, Heidelberg. Gundule, M.J., Deluca, T.H., Nordin, A., 2011. Bryophytes Bond-Lamberty, B., Gower, S.T., Amiro, B., Ewers, B.E., attenuate anthropogenic nitrogen inputs in boreal forests. 2011. Measurement and modelling of bryophyte evaporation Global Change Biology, 17, 2743–2753. in a boreal forest chronosequence. Ecohydrology, 4, 26–35. Gypser, S., Veste, M., Herppich, W., Kast, G., 2017. Linking of Buch, H.R.V., 1945. Über die Wasser- und biological soil crust wetness and ecological performance on Mineralstoffversorgung der Moose. Academic Bookstore, disturbed soils in Lower Lusatia, Germany. BES, GFÖ, Helsinki. NECOV, and EEF Joint Annual Meeting: Ecology across Carleton, T., Dunham, K., 2003. Distillation in a boreal mossy Borders. Ghent. forest floor. Canadian Journal of Forest Research, 33, 663– He, X., He, K.S., Hyvönen, J., 2016. Will bryophytes survive in 671. a warming world? Perspectives in Plant Ecology, Evolution Cornelissen, J.H., Lang, S.I., Soudzilovskaia, N.A., During, and Systematics, 19, 49–60. H.J., 2007. Comparative cryptogam ecology: a review of Hedenäs, L., 2007. Global diversity patterns among bryophyte and lichen traits that drive biogeochemistry. pleurocarpous mosses. The Bryologist, 110, 319–331. Annals of Botany, 99, 987–1001. Hillel, D., 1998. Environmental Soil Physics: Fundamentals, Dilks, T.J.K., Proctor, M.C.F., 1979. Photosynthesis, Applications, and Environmental Considerations. Elsevier. respiration and water content in bryophytes. New Leo, M., Lareo, A., Garcia-Saura, C., Hortal, J., Medina, N.G., Phytologist, 82, 97–114. 2019. BtM, a low-cost open-source datalogger to estimate Dodd, M.B., Lauenroth, W.K., 1997. The influence of soil the water content of nonvascular cryptogams. Journal of texture on the soil water dynamics and vegetation structure Visualized Experiments, 145, e58700. of a shortgrass steppe ecosystem. Plant Ecology, 133, 13–28. Li, B., Gao, J., Wang, X., Ma, L., Cui, Q., Veste, M., 2016. Einsele, G., Agster, G., 1986. Überblick zur Geologie und Effects of biological soil crusts on water infiltration and Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz evaporation Yanchi Ningxia, Maowusu Desert, China. Proctor, M.C.F., 1982. Physiological ecology: Water relations, International Journal of Sediment Research, 31, 311–323. light and temperature responses, carbon balance. In: Smith, Lindo, Z., Gonzalez, A., 2010. The bryosphere: An integral and A.J.E. (Ed.): Bryophyte Ecology. Springer, Dordrecht. influential component of the Earth’s biosphere. Ecosystems, Proctor, M.C.F., 1990. The physiological basis of bryophyte 13, 612–627. production. Botanical Journal of the Linnean Society, 104, Liu, D., She, D., 2020. Combined effects of moss crusts and 61–77. pine needles on evaporation of carbonate-derived laterite Proctor, M.C.F., 2000. The bryophyte paradox: tolerance of from karst mountainous lands. Journal of Hydrology, 586, desiccation, evasion of drought. Plant Ecology, 151, 41–49. 124859. Proctor, M.C.F., Nagy, Z., Csintalan, Z., Takács, Z., 1998. Löbs, N., Walter, D., Barbosa, C.G.G., Brill, S., Alves, R.P., Water-content components in bryophytes: Analysis of Cerqueira, G.R., de Oliveira Sá, M., de Araújo, A.C., de pressure-volume relationships. Journal of Experimental Oliveira, L.R., Ditas, F., Moran-Zuloaga, D., Pires Botany, 49, 1845–1854. Florentino, A.P., Wolff, S., Godoi, R.H.M., Kesselmeier, J., Proctor, M.C.F., Oliver, M., Wood, A., Alpert, P., Stark, L., Mota de Oliveira, S., Andreae, M.O., Pöhlker, C., Weber, B. Cleavitt, N., Mishler, B., 2007. Desiccation-tolerance in 2020. Microclimatic conditions and water content bryophytes: A review. The Bryologist, 110, 595–621. fluctuations experienced by epiphytic bryophytes in an Proctor, M.C.F., Tuba, Z., 2002. Poikilohydry and homoihydry: Amazonian rain forest. Biogeosciences, 17, 5399–5416. antithesis or spectrum of possibilities? New Phytologist, Mägdefrau, K., 1982. Life-forms of bryophytes. In: Smith, 156, 327–349. A.J.E. (Ed.): Bryophyte Ecology. Springer, Dordrecht. R Core Team, 2021. R: A language and environment for Mägdefrau, K., Wutz, A., 1951. Die Wasserkapazität der Moos- statistical computing. R Foundation for Statistical und Flechtendecke des Waldes. Veröffentlichung des Computing, Vienna, Austria. Botanischen Instituts der Forstl. Forschungsanstalt Rawls, W.J., Pachepsky, Y.A., Ritchie, J.C., Sobecki, T.M., München. Bloodworth, H., 2003. Effect of soil organic carbon on soil Medina, N., Draper, I., Lara, F., 2011. Biogeography of mosses water retention. Geoderma, 116, 61–76. and allies: Does size matter? Biogeography of microscopic Rice, S.K., Collins, D., Anderson, A.M., 2001. Functional organisms. Is everything small everywhere? Cambridge significance of variation in bryophyte canopy structure. University Press, pp. 209–233. American Journal of Botany, 88, 1568–1576. Morgan, R.P.C., 2005. Soil Erosion and Conservation. 3 edn. Rice, S.K., Gagliardi, T.A., Krasa, R.A., 2018. Canopy Blachwell Publishing, Oxford. structure affects temperature distributions and free Nakatsubo, T., 1994. The effect of growth form on the convection in moss shoot systems. American Journal of evaporation in some subalpine mosses. Ecological Research, Botany, 105, 1499–1511. 9, 245–250. Rice, S.K., Schneider, N., 2004. Cushion size, surface Nebel, M., 2001. Amblystegium serpens (Hedw.) Schimp. In: roughness, and the control of water balance and carbon flux Nebel, M., Philippi, G. (Eds.): Die Moose Baden- in the cushion moss Leucobryum glaucum (Leucobryaceae). Württembergs, Band 2: Bryophytina II, Schistostegales bis American Journal of Botany, 91, 1164–1172. Hypnobryales. Verlag Eugen Ulmer, Stuttgart, pp. 308–309. Richardson, D.H.S., 1981. The Biology of Mosses. Blackwell Nebel, M., Philippi, G., Ahrens, M., Sauer, M., Schäfer- Scientific Publications, Oxford. Verwimp, A., Schoepe, G., 2001. Die Moose Baden- Robinson, S.A., Wasley, J., Popp, M., Lovelock, C.E., 2000. Württembergs, Band 2: Bryophytina II, Schistostegales bis Desiccation tolerance of three moss species from continental Hypnobryales. Verlag Eugen Ulmer, Stuttgart. Antarctica. Australian Journal of Plant Physiology, 27, 379– Niinemets, Ü., Tobias, M., 2014. Scaling light harvesting from 388. moss “leaves” to canopies. In: Hanson, D.T., Rice, S.K. Sayer, E.J., 2006. Using experimental manipulation to assess (Eds): Photosynthesis in Bryophytes and Early Land Plants. the roles of leaf litter in the functioning of forest ecosystems. Advances in Photosynthesis and Respiration (Including Biological Reviews Cambridge Philosophical Society, 81, Bioenergy and Related Processes). Springer, Dordrecht, pp. 1–31. 151–171. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Niinemets, Ü., Tobias, M., 2019. Canopy leaf area index at its Longair, M., Pietzsch, T., Preibisch, S., Rueden, C., higher end: dissection of structural controls from leaf to Saalfeld, S., Schmid, B., Tinevez, J.Y., White, D.J., canopy scales in bryophytes. New Phytologist, 223, 118– Hartenstein, V., Eliceiri, K., Tomancak, P., Cardona, A., 133. 2012. Fiji: an open-source platform for biological-image Novák, V., Hlaváčiková, H., 2019. Applied Soil Hydrology. analysis. Nature Methods, 9, 676–682. Springer, Heidelberg, Berlin. Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Oishi, Y., 2018. Evaluation of the water-storage capacity of Image to ImageJ: 25 years of image analysis. Nature bryophytes along an altitudinal gradient from temperate Methods, 9, 671–675. forests to the Alpine zone. Forests, 9, 14. Schofield, W.B., 1981. Ecological significance of Price, A.G., Dunham, K., Carleton, T., Band, L., 1997. morphological characters in the moss gametophyte. The Variability of water fluxes through the black spruce (Picea Bryologist, 84, 149–165. mariana) canopy and feather moss (Pleurozium schreberi) Seitz, S., Nebel, M., Goebes, P., Käppeler, K., Schmidt, K., carpet in the boreal forest of Northern Manitoba. Journal of Shi, X., Song, Z., Webber, C.L., Weber, B., Scholten, T., Hydrology, 196, 310–323. 2017. Bryophyte-dominated biological soil crusts mitigate Proctor, M.C.F., 1979a. Structure and eco-physiological soil erosion in an early successional Chinese subtropical adaptation in bryophytes. In: Bryophyte Systematics. forest. Biogeosciences, 14, 5775–5788. Academic Press, London, pp. 479–509. Senf, C., Buras, A., Zang, C.S., Rammig, A., Seidl, R., 2020. Proctor, M.C.F., 1979b. Surface wax on the leaves of some Excess forest mortality is consistently linked to drought mosses. Journal of Bryology, 10, 531–538. across Europe. Nature Communications, 11, 6200. 432 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations Simon, T., 1987. The leaf-area index of three moss species CO2 pulses in biocrusted interspaces in a cool desert (Tortula ruralis, Ceratodon purpureus, and Hypnum ecosystem. Biogeochemistry, 135, 239–249. cupressiforme). In: Pócs, T., Simon, T., Tuba, Z., Podani, J. Voortman, B.R., Bartholomeus, R.P., van Bodegom, P.M., (Eds.): IAB Conference of Bryoecology. Akadémiai Kiadó, Gooren, H., van der Zee, S.E.A.T.M., Witte, J.-P.M., 2014. Budapest-Vácrátót, Hungary, pp. 699–706. Unsaturated hydraulic properties of xerophilous mosses: Slatyer, R.O., 1967. Plant-Water Relationships. Academic towards implementation of moss covered soils in Press, London, New York. hydrological models. Hydrological Processes, 28, 6251– Söderström, L., Hagborg, A., von Konrat, M., Bartholomew- 6264. Began, S., Bell, D., Briscoe, L., Brown, E., Cargill, D.C., Wang, Z., Bader, M.Y., 2018. Associations between shoot-level Costa, D.P., Crandall-Stotler, B.J., Cooper, E.D., Dauphin, water relations and photosynthetic responses to water and G., Engel, J.J., Feldberg, K., Glenny, D., Gradstein, S.R., light in 12 moss species. AoB PLANTS, 10. He, X., Heinrichs, J., Hentschel, J., Ilkiu-Borges, A.L., Weber, B., Berkemeier, T., Ruckteschler, N., Caesar, J., Heintz, Katagiri, T., Konstantinova, N.A., Larraín, J., Long, D.G., H., Ritter, H., Braß, H., Freckleton, R., 2016. Development Nebel, M., Pócs, T., Puche, F., Reiner-Drehwald, E., Renner, and calibration of a novel sensor to quantify the water M.A.M., Sass-Gyarmati, A., Schäfer-Verwimp, A., content of surface soils and biological soil crusts. Methods Moragues, J.G.S., Stotler, R.E., Sukkharak, P., Thiers, B.M., in Ecology and Evolution, 7, 14–22. Uribe, J., Váňa, J., Villarreal, J.C., Wigginton, M., Zhang, Zotz, G., Büde, B., Meyer, A., Zellner, H., Lange, O.L., 1997. L., Zhu, R.-L., 2016. World checklist of hornworts and Water relations and CO2 exchange of tropical bryophytes in liverworts. PhytoKeys, 59, 1–828. a Lower Montane Rain Forest in Panama. Botanica Acta, Soudzilovskaia, N.A., van Bodegom, P.M., Cornelissen, J.H.C., 110, 9–17. 2013. Dominant bryophyte control over high-latitude soil Zotz, G., Schweikert, A., Jetz, W., Westerman, H., 2000. Water temperature fluctuations predicted by heat transfer traits, relations and carbon gain are closely related to cushion size field moisture regime and laws of thermal insulation. in the moss Grimmia pulvinata. New Phytologist, 148, Functional Ecology, 27, 1442–1454. 59–67. Tucker, C.L., McHugh, T.A., Howell, A., Gill, R., Weber, B., Belnap, J., Grote, E., Reed, S.C., 2017. The concurrent use Received 31 March 2021 of novel soil surface microclimate measurements to evaluate Accepted 9 June 2021 SUPPORTING INFORMATION Table S1. Equations of calibration curves for studied soil substrates and moss species. The fit quality is assessed by the root mean square error (RMSE) and the determination coefficient (R ) between measured and modeled water content. Sample Calibration equation a b c d e RMSE R² –07 Angulatensandstein- y = exp(a · x) · b · x + c 0.0043 9.125e 0.264 0.027 0.928 Formation Löwenstein-Formation y = a · x + b (BWP < 1250 mV) 0.00018 0.0322 0.002 0.990 y = a · x + b (BWP > 1250 mV) 0.00215 –2.437 0.016 0.954 –06 Psilonotenton-Formation y = exp(a · x) · b · x + c 0.0038 2.055e 0.221 0.031 0.926 –08 Trossingen-Formation y = exp(a · x) · b · x + c 0.0065 3.218e 0.238 0.010 0.990 Aamblystegium serpens y = (a + b · x) –2.555 0.0045 0.703 0.979 (Lab) Brachythecium rutabulum y = a · x + b 0.0096 –0.401 0.229 0.996 Eurhynchium striatum y = a · x + b 0.0194 –0.617 0.205 0.995 Oxyrrhynchium hians y = a · x + b 0.0127 –0.414 0.200 0.993 Oxyrrhynchium hians (Lab) y = exp(a · x) · b · x + c 0.0008 0.0036 –0.057 0.133 0.998 4 3 2 –11 –08 –05 Plagiomnium undulatum y = a · x + b · x + c · x + d · x + e –3.141e 7.996e –6.066e 0.0198 –0.5402 0.223 0.994 Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz 434 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations Figs. S1–S10. Plots of calibration curves for studied soil substrates and moss species. Measured water contents for soil substrates are illustrated in brown and for moss species in green. Table S2. Maximum water storage capacity values (WSCmax) and sample sizes of the studied moss species for immersion and spray tech- –1 nique as gravimetric WSC (g g ), percentage WSC (%) and WSC per unit area (mm), ± standard error of the mean. max max max Moss species Sample WSC WSC WSC WSC WSC WSC max max max max max max –1 –1 size immersion (g g ) spray (g g ) immersion (%) spray (%) immersion (mm) spray (mm) Amblystegium 8 14.097 ± 1.28 10.082 ± 1.25 1409.668 ± 127.82 1008.176 ± 125.09 4.947 ± 0.74 3.144 ± 0.23 serpens (Lab) Brachythecium 8 11.800 ± 0.81 10.049 ± 0.66 1179.965 ± 80.52 1004.919 ± 65.74 3.152 ± 0.31 2.712 ± 0.25 rutabulum Eurhynchium 17 11.223 ± 0.62 9.629 ± 0.40 1122.260 ± 61.55 962.943 ± 39.78 3.342 ± 0.21 2.820 ± 0.18 striatum Oxyrrhynchium 7 9.686 ± 1.41 7.880 ± 0.57 968.598 ± 141.08 787.973 ± 56.90 2.094 ± 0.12 1.945 ± 0.09 hians Oxyrrhynchium 7 9.934 ± 1.24 11.038 ± 1.23 993.381 ± 123.82 1103.796 ± 122.86 2.703 ± 0.32 2.448 ± 0.21 hians (Lab) Plagiomnium 8 7.308 ± 0.80 8.146 ± 0.32 730.792 ± 79.89 814. 613 ± 31.58 1.841 ± 0.29 1.870 ± 0.13 undulatum http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Hydrology and Hydromechanics de Gruyter

Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations

Loading next page...
 
/lp/de-gruyter/water-s-path-from-moss-to-soil-a-multi-methodological-study-on-water-usVFv0nY6v
Publisher
de Gruyter
Copyright
© 2021 Sonja M. Thielen et al., published by Sciendo
ISSN
0042-790X
eISSN
1338-4333
DOI
10.2478/johh-2021-0021
Publisher site
See Article on Publisher Site

Abstract

J. Hydrol. Hydromech., 69, 2021, 4, 421–435 ©2021. This is an open access article distributed DOI: 10.2478/johh-2021-0021 under the Creative Commons Attribution ISSN 1338-4333 NonCommercial-NoDerivatives 4.0 License Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations 1, ∞ 2, ∞ 3 4 2 2, * Sonja M. Thielen , Corinna Gall , Martin Ebner , Martin Nebel , Thomas Scholten , Steffen Seitz Invertebrate Palaeontology and Palaeoclimatology, Department of Geosciences, University of Tübingen, Schnarrenbergstr. 94-96, 72076 Tübingen, Germany. Soil Science and Geomorphology, Department of Geosciences, University of Tübingen, Rümelinstr. 19-23, 72070 Tübingen, Germany. Biogeology, Department of Geosciences, University of Tübingen, Hölderlinstr. 12, 72074 Tübingen, Germany. Nees-Institute for Biodiversity of Plants, University of Bonn, Meckenheimer Allee 170, 53115 Bonn, Germany. Corresponding author. Tel.: +49 (0)7071-29-77523. E-mail: steffen.seitz@uni-tuebingen.de These authors contributed equally to this project and are considered co-first authors. Abstract: Mosses are often overlooked; however, they are important for soil-atmosphere interfaces with regard to water exchange. This study investigated the influence of moss structural traits on maximum water storage capacities (WSC ) max and evaporation rates, and species-specific effects on water absorption and evaporation patterns in moss layers, moss- soil-interfaces and soil substrates using biocrust wetness probes. Five moss species typical for Central European temperate forests were selected: field-collected Brachythecium rutabulum, Eurhynchium striatum, Oxyrrhynchium hians and Plagiomnium undulatum; and laboratory-cultivated Amblystegium serpens and Oxyrrhynchium hians. –1 –1 WSC ranged from 14.10 g g for Amblystegium serpens (Lab) to 7.31 g g for Plagiomnium undulatum when im- max –1 –1 mersed in water, and 11.04 g g for Oxyrrhynchium hians (Lab) to 7.90 g g for Oxyrrhynchium hians when sprayed, due to different morphologies depending on the growing location. Structural traits such as high leaf frequencies and small leaf areas increased WSCmax. In terms of evaporation, leaf frequency displayed a positive correlation with evapora- tion, while leaf area index showed a negative correlation. Moisture alterations during watering and desiccation were largely controlled by species/substrate-specific patterns. Generally, moss cover prevented desiccation of soil surfaces and was not a barrier to infiltration. To understand water’s path from moss to soil, this study made a first contribution. Keywords: Biological soil crusts; Bryophytes; Ecohydrology; Moss structure; Moss hydrology; Rainfall interception. INTRODUCTION special water conducting cells (hydroids), the ectohydric movement of water is through spaces between adjacent shoots, Bryophytes occur in a wide range of ecosystems, from arctic leaves, leaves and stems, leaves and rhizoids and capillary and boreal enviroments to temperate and tropical forests, dry- systems such as leaf bases, revoluted leaf margins, grooves or lands, and even deserts (Hedenäs, 2007; Lindo and Gonzalez, networks of capillary channels determined by papillae 2010; Medina et al., 2011). They often form community assem- (Giordano et al., 1993; Glime, 2017). According to Schofield blages with other organisms such as lichens, fungi, algae, cya- (1981), capillary spaces are influenced by numerous structural nobacteria and bacteria, which form what are termed biological parameters such as leaf shape, leaf arrangement, leaf orienta- soil crusts (biocrusts) (Belnap et al., 2016). With approximately tion, detailed leaf anatomy (e.g. surface ornamentation), branch 20000 species, they are the second biggest group of land plants, arrangement, nature of cortical cells, and presence of rhizoids comprising mosses, liverworts and hornworts (Frey et al., 2009; or paraphyllia. Nevertheless, there is still limited data on moss Söderström et al., 2016). Moss layers fulfill crucial functional structural traits and water relations (Elumeeva et al., 2011). roles in a variety of ecosystems regarding water and nutrient Overall, mosses achieve maximum water storage capacities of fluxes (Bond-Lamberty et al., 2011; Cornelissen et al., 2007; 108% to 2070% of their dry weight (Proctor et al., 1998), with Gundule et al., 2011) as well as soil physical properties some Sphagnum species even reaching over 5000% of dry (Soudzilovskaia et al., 2013). In contrast to vascular plants, weight (Wang and Bader, 2018). mosses do not actively regulate their water content, but are Many mosses are capable of drying out without dying, poikilohydric, meaning their internal water content is in equi- which means they can endure losing all free intracellular water librium with ambient humidity (Green and Lange, 1994). For and recover their ordinary functions afterwards, such as photo- mosses, water is primarily available via rain, dew and fog synthesizing and growing when water is available (Proctor et (Glime, 2017) and moss moisture is influenced by many fac- al., 2007). Due to their high surface to volume ratios, rapid tors, depending on the habitat as well as the species itself in drying is generally facilitated (Proctor et al., 2007). Typically, moss cells are either completely turgid or desiccated, with regard to structure and life form (Dilks and Proctor, 1979; Oishi, 2018; Proctor, 1982; Proctor, 2000; Proctor and Tuba, relatively short transitions in between (Proctor et al., 2007). 2002), i.e. the form of individual moss shoots growing together, Factors influencing this water loss by evaporation are micro- which is considered an ecologically functional unit (Bates, climatic conditions (Proctor, 1990), life form characteristics 1998; Mägdefrau, 1982). (Elumeeva et al., 2011; Mägdefrau and Wutz, 1951; Nakatsubo, Water absorption occurs mainly via the external capillaries 1994; Zotz et al., 1997) and canopy structural properties such as (ectohydric), but in some species also via internal (endohydric) surface roughness, shoot density and cushion height (Goetz and movement. While the latter is achieved cell by cell or through Price, 2015; Rice and Schneider, 2004; Rice et al., 2001, 2018). Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz As an example of cushion life forms, Zotz et al. (2000) and METHODS Rice and Schneider (2004) found that evaporation rates de- Moss and soil characteristics crease with moss cushion size. For water balance of forest ecosystems, an intact forest floor Five moss species native to Southwest Germany (Nebel et cover such as leaf litter covers or moss layers play a crucial role al., 2001) differing in origin, classification and growth form (Acharya et al., 2017; Gerrits and Savenije, 2011; Mägdefrau were chosen for the study (Table 1). Oxyrrhynchium hians and Wutz, 1951; Sayer, 2006). In mid- and high-latitude conif- (Hedw.) Loeske, Eurhynchium striatum (Hedw.) Schimp., erous forests, moss layers often form at ground level (Elbert et Plagiomnium undulatum (Hedw.) T.J.Kop. and Brachythecium al., 2012). As forest ecosystems have suffered from drought in rutabulum (Hedw.) Schimp. were collected in the field at dif- recent years (Senf et al., 2020) and mosses are also increasingly ferent sites within the Ammer and Neckar valley. Cultures of threatened by global warming (He et al., 2016), it is particularly Amblystegium serpens (Hedw.) Schimp. and Oxyrrhynchium important to investigate their hydrological effects in these envi- hians were grown in a hydraulic fluid in an in vitro environ- ronments. Previous research by Price et al. (1997) in Canadian ment by Hummel InVitro GmbH in Stuttgart, Germany. The boreal forests showed that moss layers could retain 16.8 mm of latter was selected a second time to study intraspecific differ- water, which was approximately 21% of the precipitation input. ences between field and cultivated mosses. With regard to the Furthermore, Carleton and Dunham (2003) found that mosses position of the sporophytes, all selected mosses were pleuro- in a boreal forest could not be fully hydrated by capillary water carpous (side-fruited), except P. undulatum, which was acro- movement from the forest floor or dewfall, but rather from carpous (top-fruited). vapour from the forest floor condensing on the moss surface. Soil substrates were chosen according to common growing Liu and She (2020) investigated a linear decrease of soil evapo- conditions of selected moss species and sampled from four ration with increasing moss biomass, using moss that was pre- different sites in the Schönbuch Nature Park in Southwest viously cultivated in the laboratory. Overall, the forest floor Germany. Sampling sites were located in the geological series water balance is influenced by the amounts of throughfall rain, of the Lower Jurassic, with shale clay, interstratified by beds of the processes in the moss carpet, and the processes at the moss- pyrite and fine grained sandstone, as well as in the Upper Trias- soil interface (Price et al., 1997). However, little is known sic, where claystone with fine lime nodules and fine to coarse about how much water mosses release into the atmosphere and grained sandstone is present (Einsele and Agster, 1986). The how much is transported from the soil to the moss and vice substrates varied with regard to parent material, soil texture, versa (Glime, 2017; Voortman et al., 2014). In particular, the and pH as well as the C/N ratio (Table 2). They were sampled influence of different moss species on water movement through from the topsoil to a depth of 10 cm and sieved by 6.3 mm. Be- moss layers into the soil has been largely disregarded in this low, we distinguish the substrates according to their geological context, but has in turn shown great effects on e.g. erosion formation: Angulatensandstein (AS), Psilonotenton (PT), Löwen- control (Seitz et al., 2017). stein (LS) and Trossingen (TS) (Einsele and Agster, 1986). With this study, we aim to shorten this knowledge gap in an interdisciplinary approach (cf. Liu and She (2020)). To do so, Greenhouse experiment we examined water absorption and evaporation patterns in moss-covered soil substrates typical for a Central European With a greenhouse experiment, we investigated water ab- temperate forest during and after watering. We hypothesize that: sorption patterns in moss covers and corresponding soil sub- 1. Maximum water storage capacities (WSC ) of mosses strates during and after watering. To do this, we filled the sub- max are species-specific and positively affected by their surface area. strates into infiltration boxes (40 cm × 30 cm × 15 cm) up to a 2. Differences in the temporal dynamics of water content height of 6.5 cm. Infiltration boxes are stainless steel containers during watering and subsequent desiccation depend largely on with a triangular surface runoff gutter and an outlet on the the combination of moss species and the underlying soil sub- bottom to capture percolated water. In December 2019, moss strates. species were placed onto substrate-filled infiltration boxes, To test our hypotheses, we set up a greenhouse experiment leading to 6 treatments with 2 replicates each: P. undulatum with five moss species and four soil substrates, whereby (Field) + PT, O. hians (Field) + AS, O. hians (Lab) + AS, B. artificially cultivated mosses of the same species were also rutabulum (Field) + LS, A. serpens (Lab) + LS, E. striatum + included. We used biocrust wetness probes (Weber et al., 2016) TS; yielding a total number of 12 boxes. Infiltration boxes were for high-resolution monitoring of water content in moss layers, subsequently stored in a shady place outdoors for adaptation, on the soil surface, and in a soil depth of 3 cm. Furthermore, we until we began the greenhouse experiment in July 2020. investigated the selected mosses in terms of their structural traits and their maximum water storage capacities. Table 1. Characteristics of studied moss samples. Amblystegium Brachythecium Eurhynchium Oxyrrhynchium Oxyrrhynchium Plagiomnium serpens rutabulum striatum hians hians undulatum Family Amblystegiaceae Brachytheciaceae Brachytheciaceae Brachytheciaceae Brachytheciaceae Mniaceae Origin Lab Field Field Field Lab Field Site – ruderalized fertile pinewood dry hedge – flood plain characteristics meadow understore Growth form pleurocarpous pleurocarpous pleurocarpous pleurocarpous pleurocarpous acrocarpous Sample site – Tübingen Tübingen Reusten – Pliezhausen coordinates 48.544917 N 48.546194 N 48.541665 N 48.566723 N 9.043309 E 9.036407 E 8.914316 E 9.216494 E 422 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations Table 2. Characteristics of studied soil substrates. AS PT LS TS Series Lower Jurassic Lower Jurassic Upper Triassic Upper Triassic Formation Angulatensandstein- Psilonotenton-Formation Löwenstein-Formation Trossingen-Formation Formation (AS) (PT) (LS) (TS) Parent material sandstone shale clay sandstone claystone Soil texture silt loam silty clay loam clay loam silty clay loam • sand: 7.00 % • sand: 6.88 % • sand: 25.02 % • sand: 10.78 % • silt: 67.58 % • silt: 56.28 % • silt: 42.43 % • silt: 50.83 % • clay: 25.68 % • clay: 36.93 % • clay: 32.60 • clay: 38.10 % C/N 17.54 17.36 23.12 20.05 pH 5.8 7.0 7.0 5.6 Sample site Tübingen Tübingen Tübingen Tübingen coordinates 48.553054 N 48.557425 N 48.557527 N 48.556036 N 9.119053 E 9.114462 E 9.088098 E 9.089313 To measure water content (WC), we installed three biocrust used a gravimetrical approach with a heavy-duty precision bal- wetness probes (BWP; UP GmbH, Cottbus, Germany) per ance (KERN FCB 30K1, Kern & Sohn GmbH, Balingen, Ger- infiltration box in different positions: in 3 cm soil depth, in the many), and second, we used a Thetaprobe ML2 in combination uppermost 5 mm of the soil surface and in the moss layer (Fig. with a HH2 Moisture Meter (Delta-T Devices, Cambridge, UK). 1). BWPs were specifically developed to quantify WC of soil To consider evaporation effects during the period of desicca- surfaces as well as biocrusts by deriving WC from electrical tion, we calculated the evaporation rate of this time span for all conductivity measurements; they provided reliable data in samples using the formula several experiments under different field conditions (Gypser et WC – WC 0 x al., 2017; Löbs et al., 2020; Tucker et al., 2017; Weber et al., E = , 2016). Samples were irrigated for one hour with a sprayer t – t x 0 –1 (Comfort Sensitive Plant, Gardena, Ulm, Germany) with 6 L h where WC0 is the maximum gravimetric WC in the examined of water, split into 500 mL every 5 min, corresponding to a time period, WCx is the gravimetric WC at time point x, and tx precipitation amount of 122 mm (extremely heavy rainfall and t0 are the respective time points (Robinson et al., 2000). event). All BWPs were installed underneath the centre of the sprayer, whereby we ensured that the BWP in the moss layer Laboratory BWP calibration was completely encased by moss shoots. During this watering and subsequent desiccation process in the greenhouse, the elec- To calibrate the BWP to gravimetric WC, we monitored trical conductivity of the samples was logged every 10 seconds weight loss and electrical conductivity (EC) simultaneously for for 72 hours with the BWPs connected to a GP2 Data Logger all samples under laboratory conditions for at least 65 hours (Delta-T Devices, Cambridge, UK). Additionally, air tempera- (average air temperature: 19.1 °C, sd = 1.2 °C; average relative ture and relative humidity (RH) in the greenhouse were moni- humidity (RH): 45.8%, sd = 5.9%). Samples were water satu- tored (Tinytag Plus 2 – TGP-4500, Gemini Data Loggers, rated using the immersion technique described below (in the Chichester, UK) for the same time slots. Soil WC was deter- following section). Afterwards, they were placed on a balance mined before and after watering as well as after 71 hours of (Kern EW 620-3NM, Kern & Sohn GmbH, Balingen, Germa- desiccation applying two methods: first, we substrate inside ny) and three BWPs were installed in each sample. Two sam- ples were measured in parallel, using two precision balances of the same type. BWP and weight data were recorded at an inter- val of 10 seconds, while temperature and relative humidity were logged in 5 min intervals with Tinytag Plus 2 (see above). During monitoring of weight loss, the scales generated indi- vidual error values, which required a filtering of data. Since the scales only measured stable values, we had irregular time inter- vals in the recording of weight losses. To be able to combine weight and BWP as well as RH and temperature values, we performed a linear fashion interpolation with both weight val- ues and climate measurements. As EC is affected by temperature, we conducted a tempera- ture correction and derived the WC for a specific value of the BWP as described in Weber et al. (2016). According to Slatyer (1967), the formula (WW–DW) WC = DW Fig. 1. Overview of the greenhouse experiment setup. a) Biocrust –1 was used, where WC is the gravimetric WC (g g ), WW is the wetness probe (BWP) in 3 cm soil depth, b) BWP in soil surface wet weight (g) and DW is the dry weight (g) of the soil or moss and moss cover, c) experimental setup with moss-covered soil the sample. infiltration box and sprayer installed at uniform height. Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz The last step of calibration included curve fitting, where we then drained them for 2 min, and then weighed them. We used the mean of the three BWP values and the calculated WC. decided to use these two approaches, as we observed that some We found linear relationships which can be characterized as mosses were still dry on the bottom after a rainfall event, which WC = a · EC + b. For non-linear relationships we used non- was also described in Glime (2017). Therefore, we expected linear least-square regressions expressed by the equation different mechanisms of water absorption in the two WC = exp(a · ECt) · b · ECt + c, as recommended in Weber et al. techniques, with the spray technique probably being more (2016). Furthermore, some relationships could be better de- similar to the greenhouse watering process. The soil samples scribed with the equation WC = exp(a + b · ECt). While the were placed into a tub of water until the surface was wet and moss samples could be dried from saturation to desiccation, soil afterwards we measured the wet weight. To ensure that the soil samples did not dry out completely during the laboratory cali- substrate remained in the core cutter during rewetting, we bration. Therefore, an extrapolation of data for the calibration attached a thin water permeable fleece to the bottom of the core BWP values was necessary for the soil samples. An overview cutter (Blume et al., 2011). of all calibration curves is shown in Table S1 in the supporting information. Moss structural trait measurements Maximum water storage capacity To determine the surface areas of the studied moss species, we measured the following structural traits: leaf area, leaf fre- For a detailed characterisation of moss species and adjunct quency, shoot length, length of a single component (sum of soil substrates with regard to their maximum water storage shoot length and length of attached branches), shoot density capacity (WSCmax), further laboratory experiments were con- (Table 3). We determined the surface areas of the studied spe- ducted with samples from the infiltration boxes. Therefore, we cies using the following formula, which we adapted for our detached the mosses from the soil, dried them at 30 °C in a experiment following Simon (1987), Niinemets and Tobias dehydrator (Dörrex 0075.70, Stöckli, Netstal, Switzerland) and (2014) and Niinemets and Tobias (2019): weighed the dry samples (Mettler Toledo MS603S, Mettler leaf Toledo, Columbus, USA). Soil samples were taken with A = L N bryo shoot 1 cm shoot 100 cm metal core cutters from every infiltration box, dried at 105 °C in a compartment drier and weighed in dry state. After- where A is moss surface area, L is the average length of a bryo wards both moss species and soil substrates were saturated, shoot with its attached branches, N is mean number of shoot using two different methods for the mosses: spray and immer- measured shoots, and A is mean leaf area. Leaf area index leaf sion technique. For the spray technique, we moistened the (LAI) was then calculated with the formula mosses that had been placed in a petri plate with a spray bottle from above until samples could no longer absorb water. The bryo LAI = excess water was removed with a pipette and volume was de- sample area termined with a 25 mL measuring cylinder. By weighing the spray bottle before and after spraying we estimated the amount In the first step, three circular samples with a diameter of 5.5 of water added to the mosses (in average 3.45 mm). The wet cm (sample area) were taken from each species. Moss samples mosses were weighed again with the same balance. In contrast, were then dissembled into single moss shoots. Due to the very with the immersion technique we moistened the mosses by dense structure and consequent long time duration, only half of submerging them in water for 5 min between two soil sieves the circular area of A. serpens was considered. Next, detached with 52 µm mesh size on the bottom and 250 µm on the top, shoots were scanned using a high definition flatbed scanner Table 3. Species-specific average values (± standard error of the mean) of leaf area, leaf frequency, leaf area per shoot length, shoot length, length of a single component (sum of shoot length and length of attached branches), shoot density (shoot number per ground area), total surface area, leaf area index (LAI), moss cushion height, volume and density for the studied moss species. Species Leaf Leaf fre- Leaf area Shoot Length single Shoot Total LAI Cushion Cushion Cushion area quency per shoot length component density surface height volume density 2 –1 –2 2 3 –3 (mm ) (cm ) length (cm) (cm) (n cm ) area (cm ) (cm) (cm ) (g cm ) 2 –1 (cm cm ) Amblystegium 0.104 ± 81.778 ± 0.085 ± 1.168 ± 1.764 ± 97.005 ± 346.204 14.572 1.322 ± 107.058 ± 0.026 ± serpens (Lab) 0.002 3.929 0.006 0.024 0.224 11.786 0.091 10.623 0.002 Brachythecium 1.151 ± 39.333 ± 0.452 ± 3.791 ± 8.470 ± 3.031 ± 297.076 12.504 1.536 ± 139.856 ± 0.018 ± rutabulum 0.035 4.93 0.064 0.166 0.286 0.402 0.116 19.366 0.001 Eurhynchium 0.629 ± 91.333 ± 0.574 ± 2.018 ± 7.756 ± 2.511 ± 265.672 11.182 2.119 ± 182.071 ± 0.016 ± striatum 0.013 9.541 0.06 0.129 0.656 0.496 0.092 18.683 0.002 Oxyrrhynchium 0.307 ± 69.889 ± 0.187 ± 2.524 ± 8.124 ± 4.714 ± 169.907 7.151 1.65 ± 132.174 ± 0.015 ± hians 0.006 3.545 0.008 0.129 0.702 0.712 0.13 15.278 0.002 Oxyrrhynchium 0.393 ± 55.556 ± 0.219 ± 2.180 ± 6.198 ± 10.368 ± 333.764 14.048 1.353 ± 114.336 ± 0.022 ± hians (Lab) 0.008 2.911 0.014 0.092 1.480 2.509 0.136 18.998 0.003 Plagiomnium 4.737 ± 20.111 ± 0.953 ± 3.004 ± 4.960 ± 3.087 ± 346.517 14.585 1.394 ± 100.778 ± 0.018 ± undulatum 0.129 2.6 0.121 0.129 0.571 0.827 0.08 6.649 0.001 424 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations (Epson Perfection V700 Photo, Suwa, Japan) and shoot Moss structural traits numbers of all samples were counted to determine the shoot number per unit sample area. Afterwards, if sample size A wide range of structural trait characteristics for the moss enabled it, 50 shoots were randomly chosen for length species used in this study were determined to explain moss measurements, using ImageJ versions 1.53e and Fiji 2.1.0 water relations (Table 3). The average individual leaf area of (Schindelin et al., 2012; Schneider et al., 2012). Next to shoot the studied species ranged almost fivefold from 0.104 mm² in length, we also determined the length of branches that were A. serpens (Lab) to 4.737 mm² in P. undulatum. Accordingly, attached to the measured shoots. Then, from each sample three average leaf area per shoot length varied elevenfold between 2 –1 shoots were randomly selected and all leaves were carefully 0.085 cm cm in A. serpens (Lab) to 0.953 in P. undulatum. removed along one centimeter of the shoot. The removed leaves Leaf frequency was the smallest in P. undulatum at 20.111 and were put on slides and were either scanned with the flatbed ranged up to 91.333 in E. striatum. We found the longest shoots scanner or a digital microscope (Keyence in B. rutabulum (3.79 cm on average) and the shortest shoots in VHX-7000 with dual zoom lens VH-ZST, Keyence, Osaka, A. serpens (Lab) (1.16 cm on average). After adding the length Japan). Leaf area was subsequently measured with ImageJ as of attached branches to the respective shoot length, B. rutabu- well. lum still had the longest shoots with 8.47 cm, and A. serpens Additionally, we determined the volume of the moss (Lab) had the shortest shoots with 1.764 cm. However, A. ser- cushions for all moss samples used in the WSCmax experiment. pens (Lab) had the highest shoot density (97 shoots per cm ), Therefore, we photographed all moss samples using a Nikon whereas B. rutabulum, E. striatum, O. hians and P. undulatum D5100 (Chiyoda, Japan), equipped with an AF-S DX Micro had much lower densities between 2.5 to 4.714 shoots per cm . NIKKOR 40mm f/2.8G lens to identify the individual sample Interestingly, shoot density of O. hians (Lab) was twice as high area with ImageJ. The height of the moss cushions was as for O. hians collected in nature, which might be due to miss- measured at four sites with a calliper and mean values were ing competition with other species in a laboratory setting, as calculated for every cushion. The moss cushion density was well as different light and water regimes, since moss structure derived from the quotient of dry weight and cushion volume. is highly affected by water and light availability (Mägdefrau, 1982). This raises the question of whether field-collected A. Data analysis serpens also has similarly high shoot densities as determined for A. serpens (Lab) in this study. While A. serpens (Lab) grew All analyses were conducted with R software versions 3.6.3 in dense and more voluminous lawns, A. serpens occurs more and 4.0.2 (R Core Team, 2021) on the level of individual sam- often intermingled with other species in nature. The nutrient- ples. To examine significant differences, we used one-way loving species prefers semi-shady, rather moist sites that are ANOVAs in combination with post-hoc Tukey’s HSD tests also preferred by many other species that are often more vigor- when variables showed homogeneity of variances. In other ous and thus overgrow the delicate prostrate A. serpens (Nebel, cases, we performed post-hoc Games-Howell or Wilcoxon 2001). The dense, extensive tall lawns of A. serpens (Lab) signed-rank tests. Previously, homoescedasticity was verified therefore contradict the species’ occurrence in nature and its with the Levene’s test. To test for differences of the means interspersed growth with other mosses, that can be attributed to between two samples we used Welch’s t-test. Significance was the low competitiveness of A. serpens. assessed at p < 0.05 in all cases. Compared to the other five studied species, O. hians had a Furthermore, we performed pairwise Pearson as well as low LAI of 7.151. B. rutabulum and E. striatum were similar in Spearman’s Rank correlation analyses to screen for relation- their LAI of 12.504 and 11.182, respectively, and highest LAI ships between WSC as well as evaporation rates of the stud- values were determined for A. serpens (Lab) (14.572), O. hians max ied samples and parameters of sample characteristics. In ad- (Lab) (14.048) and P. undulatum (14.585). Interestingly, P. vance of all analyses, we used the Shapiro-Wilk Test to exam- undulatum and the two lab-grown mosses are very different in ine the samples for normal distribution. Additionally, general- terms of leaf area, leaf frequency and shoot density, but all have ized additive models (GAM) with restricted maximum likeli- similar LAI values. Considering the moss cushion density, A. hood and smoothing parameters selected by an unbiased risk serpens (Lab) was significantly denser than E. striatum (p < estimator (UBRE) criterion were performed to assess the effect 0.001), O. hians (p < 0.001) and P. undulatum (p < 0.01). Fur- of soil substrate or moss species characteristics on WSCmax. thermore, we found significant differences in regard to moss Firstly, we fitted moss WSCmax from the spray and immersion cushion density between O. hians (Lab) and E. striatum (p < techniques against mean shoot number, mean leaf surface area, 0.01), O. hians (Lab) and O. hians (p < 0.05), B. rutabulum and LAI, moss cushion height as well as moss cushion density. E. striatum (p < 0.05) and E. striatum and P. undulatum (p < Secondly, WSC of soil substrates were fitted against soil 0.05). max bulk density, sand, silt and clay contents as well as total carbon and nitrogen content. Maximum water storage capacity RESULTS AND DISCUSSION Mean values of WSC from the immersion technique (rep- max –1 resenting complete soaking) varied between 14.10 g g for A. –1 In order to discuss and answer the hypotheses presented, we serpens (Lab) and 7.31 g g for P. undulatum, with the differ- first analyzed the differences in structural traits of the studied ence being highly significant (p < 0.001) (Fig. 2 and Table S2 moss species and investigated their relationship with WSCmax. in the supporting information). Further significant differences –1 As we assumed that the temporal progression of WC in the were found between E. striatum (11.22 g g ) and P. undulatum –1 greenhouse experiment could be explained by the structural as well as between B. rutabulum (11.80 g g ) and P. undulatum traits of moss species, we further examined whether our sam- (p < 0.05). Thus, with regard to the WSCmax, there were strong ples showed similar patterns of properties in the different ex- differences between the mosses with different growth forms, periments. but none within the group of pleurocarpous mosses. The fact Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz –1 Fig. 2. Maximum water storage capacity (g g ) of treatments (moss species + soil substrate). For moss species both spray and immersion technique are illustrated. Crosses represent mean values and lines within boxplots median values. The bottom and top of the box represent the first and third quartiles, and whiskers extend up to 1.5 times the interquartile range (IQR) of the data. Outliers are defined as more than 1.5 times the IQR and are displayed as points. that P. undulatum absorbed comparatively less water could be importance to the WSC . Such parameters are assumed to be, max explained by its endohydric water transport, and many acrocar- for example, the capillary spaces of mosses, which are very pous mosses are endohydric (Richardson, 1981). Since the difficult to quantify and are diverse and often complex (Proctor, surface of endohydric mosses comprises a water-resistant cuti- 1982). According to Proctor (1982), capillary conducting sys- cle with often waxy layers (Buch, 1945; Proctor, 1979a; tems such as spaces between overlapping leaves, between 1979b), water absorption through their leaves is inhibited shoots, in sheathing leaf bases or amongst rhizoid tomentum (Glime, 2017). However, as we only measured one acrocarpous and paraphyllia can be 10–100 µm large. In addition, interspac- moss, this finding requires further investigation. es of a few µm can be found in interstices between papillae as Although the most significant difference in WSCmax was well as in furrows between plicae and ridges on leaves and shown between the visibly densest and loosest growing moss stems (Proctor, 1982). In this context, the 3D structure of the species, this relationship could not be substantiated by the mosses, e.g. the branching of the shoots, the shape of the leaves surveyed traits for surface area and cushion characteristics. and the position of the leaves in relation to the stems, potential- WSC was not affected by total surface area or LAI. Further- ly plays an important role for capillarity of bryophytes max more, neither height of the moss cushions, nor volume or densi- (Giordano et al., 1993; Schofield, 1981). ty correlated individually with WSC . The combination of In contrast to the immersion technique, the range of mean max leaf area and leaf frequency seemed to have a higher influence values of WSC for the spray technique, which was intended max on WSC : with a small leaf area (Spearman’s correlation to simulate moistening of mosses by a rainfall event, was con- max rho = –0.30, p < 0.05) and high leaf frequency (Spearman’s siderably smaller (Fig. 2). Here, we found a variation of 11.04 –1 –1 correlation rho = 0.32, p < 0.05), the WSCmax increased. Shoot g g for O. hians (Lab) to 7.90 g g for O. hians from the field. density might be another influencing factor, but due to small However, we could not find any significant differences between sample size further studies are recommended. In this context, species or significant correlations between the WSCmax and the Voortman et al. (2014) also discussed that capillary spaces ascertained individual moss structure parameters, and the ad- between moss leaves and branches might be more relevant for junct GAM could explain 46.5% of the deviations. The greatest water retention than those between moss shoots. For Sphagnum influence was due to moss height (p < 0.01), with LAI having a species, Bengtsson et al. (2020) also found a high influence of smaller effect (p < 0.05). Interestingly, the greatest difference leaf traits on water retention. in WSC was discovered within the same species, O. hians. max Calculated in a GAM explaining 54.1% of the deviance, Although they belong to the same species, O. hians collected in moss cushion density highly influenced WSC (p < 0.001), the field and O. hians grown in the laboratory displayed strong max while the effects of mean leaf area (p < 0.01) and mean shoot differences in structure. While O. hians grows as loose lawn density (p < 0.05) were smaller, but also significant. Therefore, in the field, the laboratory variety forms very dense moss we assume that additional parameters must be also of great cushions, which is also reflected in the higher shoot density 426 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations (O. hians: 4.714 shoots per cm and O. hians (Lab): 10.368 with bulk density (Pearson’s correlation r = –0.70, t = –5.94, p shoots per cm ), and the larger total surface area (O. hians: < 0.001) and C/N ratio (Spearman’s correlation rho = –0.62, p 2 2 169.907 cm and O. hians (Lab): 333.764 cm ). This finding < 0.001). Additionally, we tested for a joint impact on soil indicates that the WSC of mosses is dependent on life form. WSC using a GAM with soil bulk density, sand, silt and clay max max In a further chain of thought, this also implies that single spe- contents as well as total carbon and nitrogen content as fixed cies can obtain more advantageous properties through laborato- effects and were able to explain 84.7% of the deviance with this ry cultivation, e.g. for erosion control. model. The results also showed a high relevance of bulk soil Overall, we suppose that for both the immersion technique density as well as total carbon content (p < 0.001), which is and the spray technique, the capillary spaces between moss consistent with the results of the individually tested correlations shoots as well as between leaves and shoots are more important and an influence of the clay content (p < 0.01). These relation- for WSC than surface parameters such as LAI or total sur- ships are also reported in other studies (Gong et al., 2003; max face area. Finally, it can be concluded that a further develop- Franzluebbers, 2002; Novák and Hlaváčiková, 2019; Rawls et ment and standardization of the spray technique is required to al., 2003). be able to gather more reliable data on this important moss characteristic. Greenhouse experiment Regarding the soil substrates, WSC values varied on av- Watering process max –1 –1 erage between 0.46 g g for PT and 0.36 g g for TS, which is 30 times less compared to the WSCmax of the mosses (Fig. 3). Focusing on the 60 minutes of watering, we observed clear Within soil substrates we found highly significant differences differences in WC of different moss species, regarding tem- between PT and TS, PT and LS as well as AS and TS poral progression as well as the level of WC achieved (Fig. 3). (p < 0.001) and a significant difference between AS and TS At the beginning of the watering, all moss species were desic- (p < 0.05). On one hand, these differences can be explained by cated, so that the WC initially increased until an equilibrium soil texture, as there is a negative relationship with sand content was reached. Moss species were classified in terms of WC in –1 equilibrium: (a) low WC (0–5 g g (Spearman’s correlation rho = –0.62, p < 0.001) and a positive ) for A. serpens (Lab) and P. –1 correlation with silt content (Spearman’s correlation rho = 0.52, undulatum, (b) medium WC (5–10 g g ) for B. rutabulum, O. –1 p < 0.001), while the clay content seemed to be of rather minor hians and O. hians (Lab), (c) high WC (10–15 g g ) for E. importance for WSC (Spearman’s correlation rho = –0.40, p striatum. This classification shows the possibility of distin- max < 0.01). On the other hand, we revealed a negative correlation guishing between moss species based on the BWP response. -1 Fig. 3. Temporal progression of water content values (g g ) of treatments during watering in the greenhouse experiment. Replicate measurements are labelled with A and B for every biocrust wetness probe (BWP) location (moss cover, soil surface, 3 cm soil depth). Plotted are half-minute values. Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz Surprisingly, A. serpens (Lab) and P. undulatum both reached a Compared to the mosses, the soil substrates showed a much low WC during irrigation, although they are quite different re- lower WC during the 60 minutes of watering, which is true for garding their structural traits. While A. serpens (Lab) forms very both the surface and 3 cm soil depth (Fig. 3). Overall, mosses –1 dense moss cushions (shoot density: 97.005 ± 11.786 shoots per adjusted their equilibrium in the range between 2.5–15.0 g g 2 –1 cm ), P. undulatum is more likely to grow single shoots (shoot of WC, while soil substrates varied between 0.15–0.35 g g . density: 3.087 ± 0.827 shoots per cm ). Although O. hians and O. The fact that mosses can absorb more water than soil substrates hians (Lab) were both assigned to medium WC, we recognized a could be attributed to a larger surface area of mosses. Addition- –1 distinct difference, with O. hians tending to weigh 10 g g and O. ally, capillary effects in mosses might contribute to higher –1 hians (Lab) tending to weigh 5 g g . Since O. hians (Lab) grows water absorption rates compared to soil substrates. considerably denser than O. hians with a shoot density twice as Since the soil surfaces were not completely dried out at the high (O. hians (Lab) = 10.368 ± 2.509 shoots per cm , O. hians = beginning of the experiment, they showed a relatively high 4.714 ± 0.712 shoots per cm ) and a higher cushion density (O. starting value of WC in comparison with the later reached hians (Lab) = 0.022 ± 0.003 shoots per cm , O. hians = 0.015 ± equilibrium. The temporal progression of WC on the soil sur- 0.002 shoots per cm ), we expected that O. hians (Lab) would face started with higher values at the beginning of watering and also absorb more water during watering. The fact that this ex- slightly decreased over time. Regarding infiltration into the soil pectation was not fulfilled could be attributed to O. hians hav- surface, it appeared that water had initially accumulated on the ing a comparatively high leaf frequency with small leaf area, surface, causing the high WC. which had already been highlighted as important factors for When considering WC at 3 cm soil depth, temporal progres- water absorption in previous chapters. sion of WC was almost steady, which was also due to the al- Furthermore, almost all moss species showed a certain varia- ready wet soil substrate at the beginning of the experiment. For tion in WC at equilibrium within replicate measurements, illus- two substrates (AS and PT) we observed an increase of WC trating a great heterogeneity within species. Overall, we noticed during the first 10 minutes of irrigation, indicating percolation that the variations between replicate measurements were small- of water through the substrate. Additionally, WC tended to be er for denser moss cushions than for looser ones, with P. undu- lower at 3 cm soil depth than on the soil surface during irriga- latum being an exception in this case. This could be attributed tion. Overall, with respect to the temporal progression of WC to the fact that denser mosses establish better contact with the values on soil surface and in 3 cm soil depth, we generally sensor without forming air spaces (Löbs et al., 2020). found substrate-specific coherences regarding the level of WC Some moss species demonstrated a more pronounced re- achieved. sponse to the watering pulses than others. This might also be Furthermore, we expected that the soil substrates show a related to denser moss cushions with less air-filled interstitial similar response due to WSC in the lab and in the green- max spaces (Löbs et al., 2020), as it was the case for A. serpens (Lab) house experiment. However, the WC after watering in the and O. hians (Lab), which both form the densest cushions. To greenhouse, which we expected to be the maximum WC examine moss intraspecific differences regarding water absorp- reached in the greenhouse (means of WC for all values between th th tion in detail, higher replication is necessary in future studies. 60 and 65 minute), were lower than the WSCmax measured in Because of the water volume applied to the samples in the the lab, which was true for every substrate both for surface as greenhouse, we speculated that the moss species would reach well as in 3 cm soil depth. For example, PT achieved a WSCmax –1 –1 their WSCmax within the watering time in the greenhouse, espe- of 0.46 g g and only showed a WC of 0.31 g g on the surface –1 cially when compared with the achieved WSCmax using the and 0.27 g g in the soil after one hour of watering in the spray technique. To go into more detail, we compared the WC greenhouse, which means a deviance of 32.61%. In compari- values directly after watering (means of WC for all values son, LS reached only 50% of the WSC under A. serpens max th th –1 –1 between 60 and 65 minute) with the WSC determined in (Lab) (WC after watering = 0.19 g g , WSC = 0.46 g g ) max max the lab. For most of the species the WC after watering was and 52% under B. rutabulum on the surface (WC after watering –1 considerably lower than the WSC , for both spray and immer- = 0.18 g g ), WC values in 3 cm soil depth were even lower max –1 sion technique. As an example, the maximum WC for A. ser- (WC after watering (A. serpens (Lab)) = 0.18 g g ; WC after –1 pens using the immersion technique was 5 times higher than the watering (B. rutabulum) = 0.17 g g ). Altogether, soil sub- –1 WC after watering (WSCmax (immersion) = 14.10 g g , SE = strates did not show the same patterns of water absorption in –1 1.28, WC after watering = 2.63 g g , SE = 0.02), while the the lab as in the greenhouse. spray technique showed almost a fourfold difference (WSCmax –1 (spray) = 10.10 g g , SE = 1.25). Additionally, we found an Desiccation process almost fivefold difference from the immersion technique, re- spective fourfold difference from the spray technique, and During the subsequent desiccation process of 71 hours, higher WSCmax compared to the WC after watering in P. undu- moisture in the moss layers generally decreased, while moisture –1 latum (WSC (immersion) = 7.31 g g , SE = 0.80, WSC at the soil substrate surface as well as in 3 cm soil substrate max max –1 –1 (spray) = 8.15 g g , SE = 0.32, WC after watering = 1.76 g g , depth remained at the same levels (Fig. 4). However, moss SE = 0.01). Based on these results, no clear patterns are species differed in maximum WC, evaporation rates and their discernible that would explain the different intraspecific mech- responses to climatic changes in the greenhouse. Sample repli- anisms of water absorption comparing greenhouse and labora- cates slightly differed from each other in regard to WC values, tory experiments. Above all, it was very surprising that espe- but generally showed comparable patterns. We observed the cially the denser mosses, most notably the lab-grown mosses, highest WC values directly after watering in E. striatum with a –1 did not absorb much water during the greenhouse experiment. mean WC of almost 15 g g , while mean WC of B. rutabulum, –1 In general, we can deduce that the mosses are not a barrier to O. hians and O. hians (Lab) ranged between 5–10 g g , and infiltration in case of high precipitation rates, as also reported in mean WC of A. serpens (Lab) and P. undulatum did not exceed –1 Li et al. (2016). A new observation of our study is that the 5 g g . The low WC of P. undulatum might be related to its mosses growing on the soil do not store much of the applied delicate and loose structure with a low leaf frequency and large water themselves, but pass it on to the soil. leaf areas, and leaves that stand off the shoot. Especially 428 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations –1 Fig. 4. Temporal progression of water content values (g g ) of treatments during 71 h of desiccation in the greenhouse experiment. Repli- cate measurements are labelled with A and B for every biocrust wetness probe (BWP) location (moss cover, soil surface, 3 cm soil depth). Mean temperature and mean relative humidity ± standard deviation: Amblystegium serpens (Lab) + Löwenstein-Formation (LS) 25.93 ± 6.13 °C, 42.67 ± 14.39%; Brachythecium rutabulum + Löwenstein-Formation (LS) 26.35 ± 5.38 °C, 49.45 ± 15.22%; Eurhynchium stria- tum + Trossingen-Formation (TS) 24.70 ± 5.78 °C, 46.31 ± 16.15%; Oxyrrhynchium hians + Angulatensandstein-Formation (AS) 20.30 ± 3.89 °C, 64.72 ± 18.45%; Oxyrrhynchium hians (Lab) + Angulatensandstein-Formation (AS) 23.10 ± 6.07 °C, 53.37 ± 18.38%; Plagiomni- um undulatum + Psilonotenton-Formation (PT) 20.96 ± 4.31 °C, 59.92 ± 15.35%. Plotted are hourly values. –1 compared to a branched structure with high leaf frequencies ably smaller (0.023 and 0.012 g h ). A group with slightly and densely attached leaves, few capillary spaces for water higher evaporation rates consisted of A. serpens (Lab) (0.056 –1 –1 storage are formed in P. undulatum (Mägdefrau and Wutz, and 0.03 g h ), B. rutabulum (0.046 and 0.055 g h ), O. hians –1 1951). Furthermore, leaf surfaces of mosses from the Mniaceae (Lab) (0.057 and 0.078 g h ) and O. hians (0.06 and 0.093 –1 family often have a water-resistant cuticle, reducing their g h ). We found a positive relationship between leaf frequency ability to absorb water via the leaves (Glime, 2017; Proctor, and evaporation rate (Spearman’s correlation rho = 0.832, P < 2000). Additionally, we observed that leaves and stems of P. 0.001). LAI, however, correlated negatively with evaporation undulatum were twisting and curling during the desiccation rate (Spearman’s correlation rho = –0.78, P < 0.001); this was process, which might result in altered measurement conditions congruent with our expectations of lower evaporation rates for for the sensor. Clipping the sensor to moss stems of such moss species with a high LAI, which, as a product of different species as P. undulatum, as proposed in Leo et al. (2019), structural traits, makes the formation of a multitude of capillary would be interesting to compare with BWP response in future spaces for water storage in different hierarchical levels (leaf, studies. Nevertheless, the BWP used in this study proved to be shoot, and colony) more likely, overall resulting in wetter moss successful in all moss species, as also confirmed in Löbs et al. cushions and lower evaporation rates, as also described in (2020). Elumeeva et al. (2011). A. serpens (Lab) had dried out after 30 hours, whereas the WC in moss species showed diel fluctuations, albeit to other species generally remained moist longer than 40 hours, different degrees. Desiccation periods clearly aligned with and did not desiccate completely during the measurement. A declining RH and rising temperatures in E. striatum, O. hians more stabilized, steady evaporation was observed in B. rutabu- (Lab), and to a smaller degree in P. undulatum, A. serpens lum, O. hians, O. hians (Lab) and P. undulatum. Evaporation (Lab), B. rutabulum and O. hians. Comparably high RH and rates calculated for the measurement period corresponded to low temperatures contributed to the quite stable WC of O. hians maximum WC: E. striatum with the highest maximum WC throughout the measurement and to the fact that the moss did after watering also had the highest evaporation rates (0.181– not dry completely. We observed slight reactions of WC –1 0.197 g h ). Evaporation rates for P. undulatum were consider- towards RH changes in all samples, confirming that mosses Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz reacted to increasing RH and could absorb water under CONCLUSIONS AND OUTLOOK conditions with high RH, as also described in Löbs et al. (2020). This study found that five moss species from Central Euro- Climatic conditions cannot explain intraspecific variation of pean temperate forests can exhibit different water absorption WC, since the replicates were measured in parallel at the same and evaporation patterns in response to rainfall. In some cases, time. A possible explanation could be that moss structure at the the target moss species also showed significant intraspecific sensor locations differed in regard to surface roughness, alter- variability in rainwater interception. With regard to our hypoth- ing boundary layer resistance and thus resulting in different eses, the following conclusions were drawn: evaporation velocities (Proctor, 1982). Further experiments in a 1. Contradictory to our hypothesis, total surface area did climate-controlled environment with closer control and manipu- not affect maximum water storage capacity (WSCmax). Results lation possibilities could determine if moss reactions are spe- further indicate that a combination of structural traits (high cies-specific. shoot density, high leaf frequency, and low leaf area) may The different soil substrates had slightly different mean WC increase WSC during immersion. Generalized additive max –1 –1 values in 3 cm depth: LS 0.16–0.18 g g , TS 0.24 g g , AS models (GAM) revealed that cushion density also can influence –1 –1 0.28 g g and PT 0.24 g g . In LS, a slight reaction to rising WSC . A combination of different structural traits tested in a max RH (due to night-day-shifts) was recognizable, and LS did not GAM showed that WSC determined using the spray max desiccate, despite high temperatures above 40 °C during the technique was affected by leaf area index (LAI) and moss measurement period. We assume that the moss cover prevented height. Overall, soil substrates absorbed around 30 times less desiccation of the substrate, but it remains unclear whether the water compared to mosses and an effect of bulk density, grain substrate receives water from the moss cushion itself or plainly size distribution and total carbon content on WSCmax was from RH. For low precipitation rates, prevention of soil evapo- found. ration from moss-dominated biocrusts was also reported in Li et 2. Both moss species and soil substrates showed al. (2016). species/substrate-specific patterns in regard to changes of WC at the soil surface fluctuated diurnally depending on RH moisture during watering as well as desiccation. Since soil as also described in Tucker et al. (2017), especially in AS and substrates did not desiccate despite high temperatures, yet water PT and less pronounced in LS. Moreover, we found that content at the surface responded to relative humidity changes, oscillations related to RH were visible at the soil surface but not we hypothesize that the moss cover prevented desiccation in 3 cm soil depth, which showed that fine pores at the surface without sealing the soil. Because the humidity-induced were capable of adsorbing water out of the air (Agam and fluctuations varied depending on the density of the moss cover, Berliner, 2006; Hillel, 1998). So even dense moss cushions we further hypothesize that mosses mitigate soil evaporation. were not completely sealing the soil surface and there was no Among moss species, differences were also observed between full barrier by bryophytes. However, since the RH-induced replicates, primarily related to the moistening until an fluctuations varied depending on the density of the moss cover, equilibrium in water content was reached, as well as in the i.e. the most pronounced reactions were found in the loosest process of desiccation. Similar WSCmax values (to immersion moss cover P. undulatum, we assume that mosses mitigate soil and spray) were not achieved in greenhouse experiments during evaporation. watering, indicating different mechanisms of water absorption Generally, WC at the soil surface was higher than in 3 cm for both soil substrates and moss species, which could not be depth during desiccation. This could be ascribed to the fact that explained by clear patterns. In general, we can deduce that the the soil surface had a finer texture due to clogging of the pores mosses growing on the soil may not store much of the applied as an influence of splash effects (Morgan, 2005), which allows water themselves, but pass it on to the soil. Leaf frequency for a higher WC (Dodd and Lauenroth, 1997). A further influ- correlated positively with evaporation rates, while LAI showed encing factor to explain this observation might be the initial soil a negative relationship with evaporation rates. WC. As we measured a high soil WC before watering, the Although not explicitly mentioned in our hypotheses, the matrix potential is reduced, resulting in a lower and less deep results underscore that some species can develop different infiltration (Novák and Hlaváčiková, 2019). morphologies due to different growing locations (field vs. la- Differences between WC values of surface and 3 cm depth boratory). This can lead to a heterogeneous expression of the depended on the substrate: for LS, the values were very similar, same traits and raises the question of whether beneficial traits but especially for PT, WC values at substrate surface were can be conferred to individual species by laboratory cultivation, higher than in 3 cm depth by a factor of 1.4 to 2.3. In AS, there e.g. for erosion control. Thus, the interplay of individual moss was either an influence by the moss cover, or by the climatic structure traits appears to be very complex, such that further conditions during the measurement: AS covered with O. hians detailed investigations especially on the 3D structure of indi- showed a smaller difference between surface and soil WC and vidual species are urgently needed. In this context, more infor- not very pronounced oscillations with RH. In contrast, AS mation on moss capillary spaces would help to achieve a higher covered with O. hians (Lab) displayed strong day-night level of accuracy regarding the mechanisms of water absorption oscillations and WC values during nights were up to 1.5-times in mosses. It should be noted that the methodology also needs higher in the surface than in 3 cm depth. Since RH remained further improvement and the exact determination of individual above 50% after 20 hours during the measurement of AS with species effects can be seen as non-trivial. O. hians, but dropped from 75% in the nights to 25% RH Considering that the methodology has proven to be sound, during the days in the measurement of AS with O. hians (Lab), the full significance of the current results in this study needs to we cannot exclude a strong influence of these fluctuations on be confirmed in a larger experimental setup. Further research is the different oscillation patterns in the AS measurements. To required to understand the details of how different moss species determine the effect of moss layers itself on soil substrate and soil substrates interact regarding water absorption and moistness and evaporation, an experiment with different moss evaporation. A multi-method approach to measure water con- species on similar substrates and control samples without moss tent in different layers is recommended, using biocrust wetness is necessary. probes as well as clip sensors for the moss cover as introduced 430 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations by Leo et al. (2019). This approach should be combined with Morphologie des Schönbuchs. In: Einsele, G. (Ed.): Das the use of a climate-regulated greenhouse and expanded to landschaftsökologische Forschungsprojekt Naturpark include control samples without moss cover and large number Schönbuch: Wasser- und Stoffhaushalt, Bio-, Geo- und of replicates in order to cover the existing complexity as well as Forstwirtschaftliche Studien in Südwestdeutschland. VCH possible. This complexity is also the major challenge in the Verlagsfesellschaft, Weinheim. investigation of "water’s path from moss to soil", to the under- Elbert, W., Weber, B., Burrows, S., Steinkamp, J., Büdel, B., standing of which this study has made a further contribution. Andreae, M.O., Pöschl, U., 2012. Contribution of cryptogamic covers to the global cycles of carbon and Acknowledgements. We thank Michael Sauer for his assistance nitrogen. Nature Geoscience, 5, 459–462. and expertise during moss sample collection, Lena Grabherr Elumeeva, T.G., Soudzilovskaia, N.A., During, H.J., and Sarah Fodor for their support in measuring moss structural Cornelissen, J.H., 2011. The importance of colony structure parameters, Daniel Schwindt for helpful comments and Mat- versus shoot morphology for the water balance of 22 thew Hughes for improving the language of the manuscript. We subarctic bryophyte species. Journal of Vegetation Science, are also grateful to the Plant Ecology group of the University of 22, 152–164. Tübingen for the space provided in their greenhouse. We thank Franzluebbers, A.J., 2002. Water infiltration and soil structure three anonymous reviewers for their constructive comments. related to organic matter and its stratification with depth. This study was funded by the German Research Foundation Soil and Tillage Research, 66, 197–205. (DFG SE 2767/2-1 “MesiCrust”) and the Federal Ministry for Frey, W., Stech, M., Fischer, E., 2009. Syllabus of Plant Economic Affairs and Energy (AiF 19808 N). Families - Part 3: Bryophytes and Seedless Vascular Plants. Borntraeger, Berlin, Stuttgart. REFERENCES Gerrits, A.M.J., Savenije, H.H.G., 2011. Forest floor interception. In: Levia, D.F., Carlyle-Moses, D., Tanaka, T. Acharya, B.S., Stebler, E., Zou, C.B., 2017. Monitoring litter (Eds.): Forest Hydrology and Biogeochemistry: Synthesis of interception of rainfall using leaf wetness sensor under Past Research and Future Directions. Springer Netherlands, controlled and field conditions. Hydrological Processes, 31, Dordrecht, pp. 445–454. 240–249. Giordano, S., Colacino, C., Spagnuolo, V., Basile, A., Esposito, Agam, N., Berliner, P.R., 2006. Dew formation and water A., Castaldo-Cobianchi, R., 1993. Morphological adaptation vapor adsorption in semi-arid environments – A review. to water uptake and transport in the poikilohydric moss Journal of Arid Environments, 65, 572–590. Tortula ruralis. Giornale Botanico Italiano, 127, 1123–1132. Bates, J.W., 1998. Is 'life-form' a useful concept in bryophyte Glime, J.M., 2017. Volume 1: Physiological Ecology. ecology? Oikos, 82, 223–237. Bryophyte Ecology. Belnap, J., Weber, B., Büdel, B., 2016. Biological soil crusts as Goetz, J.D., Price, J.S., 2015. Role of morphological structure an organizing principle in drylands. In: Weber, B., Büdel, and layering of Sphagnum and Tomenthypnum mosses on B., Belnap, J. (Eds.): Biological Soil Crusts: An Organizing moss productivity and evaporation rates. Canadian Journal Principle in Drylands. Springer International Publishing, of Soil Science, 95, 109–124. Cham, pp. 3–13. Gong, Y., Cao, Q., Sun, Z., 2003. The effects of soil bulk Bengtsson, F., Granath, G., Cronberg, N., Rydin, H., 2020. density, clay content and temperature on soil water content Mechanisms behind species-specific water economy measurement using time-domain reflectometry. responses to water level drawdown in peat mosses. Annals Hydrological Processes, 17, 3601–3614. of Botany, 126, 219–230. Green, T.G.A., Lange, O.L., 1994. Photosynthesis in Blume, H.-P., Stahr, K., Leinweber, P., 2011. Bodenkundliches poikilohydric plants: a comparison of lichens and Praktikum: eine Einführung in pedologisches Arbeiten für bryophytes. In: Schulze, E.-D., Caldwell, M.M. (Eds.): Ökologen, insbesondere Land- und Forstwirte, und für Ecophysiology of Photosynthesis. Springer, New York, pp. Geowissenschaftler. 3. neubearbeitete Auflage edn. 319–341. Spektrum Akademischer Verlag, Heidelberg. Gundule, M.J., Deluca, T.H., Nordin, A., 2011. Bryophytes Bond-Lamberty, B., Gower, S.T., Amiro, B., Ewers, B.E., attenuate anthropogenic nitrogen inputs in boreal forests. 2011. Measurement and modelling of bryophyte evaporation Global Change Biology, 17, 2743–2753. in a boreal forest chronosequence. Ecohydrology, 4, 26–35. Gypser, S., Veste, M., Herppich, W., Kast, G., 2017. Linking of Buch, H.R.V., 1945. Über die Wasser- und biological soil crust wetness and ecological performance on Mineralstoffversorgung der Moose. Academic Bookstore, disturbed soils in Lower Lusatia, Germany. BES, GFÖ, Helsinki. NECOV, and EEF Joint Annual Meeting: Ecology across Carleton, T., Dunham, K., 2003. Distillation in a boreal mossy Borders. Ghent. forest floor. Canadian Journal of Forest Research, 33, 663– He, X., He, K.S., Hyvönen, J., 2016. Will bryophytes survive in 671. a warming world? Perspectives in Plant Ecology, Evolution Cornelissen, J.H., Lang, S.I., Soudzilovskaia, N.A., During, and Systematics, 19, 49–60. H.J., 2007. Comparative cryptogam ecology: a review of Hedenäs, L., 2007. Global diversity patterns among bryophyte and lichen traits that drive biogeochemistry. pleurocarpous mosses. The Bryologist, 110, 319–331. Annals of Botany, 99, 987–1001. Hillel, D., 1998. Environmental Soil Physics: Fundamentals, Dilks, T.J.K., Proctor, M.C.F., 1979. Photosynthesis, Applications, and Environmental Considerations. Elsevier. respiration and water content in bryophytes. New Leo, M., Lareo, A., Garcia-Saura, C., Hortal, J., Medina, N.G., Phytologist, 82, 97–114. 2019. BtM, a low-cost open-source datalogger to estimate Dodd, M.B., Lauenroth, W.K., 1997. The influence of soil the water content of nonvascular cryptogams. Journal of texture on the soil water dynamics and vegetation structure Visualized Experiments, 145, e58700. of a shortgrass steppe ecosystem. Plant Ecology, 133, 13–28. Li, B., Gao, J., Wang, X., Ma, L., Cui, Q., Veste, M., 2016. Einsele, G., Agster, G., 1986. Überblick zur Geologie und Effects of biological soil crusts on water infiltration and Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz evaporation Yanchi Ningxia, Maowusu Desert, China. Proctor, M.C.F., 1982. Physiological ecology: Water relations, International Journal of Sediment Research, 31, 311–323. light and temperature responses, carbon balance. In: Smith, Lindo, Z., Gonzalez, A., 2010. The bryosphere: An integral and A.J.E. (Ed.): Bryophyte Ecology. Springer, Dordrecht. influential component of the Earth’s biosphere. Ecosystems, Proctor, M.C.F., 1990. The physiological basis of bryophyte 13, 612–627. production. Botanical Journal of the Linnean Society, 104, Liu, D., She, D., 2020. Combined effects of moss crusts and 61–77. pine needles on evaporation of carbonate-derived laterite Proctor, M.C.F., 2000. The bryophyte paradox: tolerance of from karst mountainous lands. Journal of Hydrology, 586, desiccation, evasion of drought. Plant Ecology, 151, 41–49. 124859. Proctor, M.C.F., Nagy, Z., Csintalan, Z., Takács, Z., 1998. Löbs, N., Walter, D., Barbosa, C.G.G., Brill, S., Alves, R.P., Water-content components in bryophytes: Analysis of Cerqueira, G.R., de Oliveira Sá, M., de Araújo, A.C., de pressure-volume relationships. Journal of Experimental Oliveira, L.R., Ditas, F., Moran-Zuloaga, D., Pires Botany, 49, 1845–1854. Florentino, A.P., Wolff, S., Godoi, R.H.M., Kesselmeier, J., Proctor, M.C.F., Oliver, M., Wood, A., Alpert, P., Stark, L., Mota de Oliveira, S., Andreae, M.O., Pöhlker, C., Weber, B. Cleavitt, N., Mishler, B., 2007. Desiccation-tolerance in 2020. Microclimatic conditions and water content bryophytes: A review. The Bryologist, 110, 595–621. fluctuations experienced by epiphytic bryophytes in an Proctor, M.C.F., Tuba, Z., 2002. Poikilohydry and homoihydry: Amazonian rain forest. Biogeosciences, 17, 5399–5416. antithesis or spectrum of possibilities? New Phytologist, Mägdefrau, K., 1982. Life-forms of bryophytes. In: Smith, 156, 327–349. A.J.E. (Ed.): Bryophyte Ecology. Springer, Dordrecht. R Core Team, 2021. R: A language and environment for Mägdefrau, K., Wutz, A., 1951. Die Wasserkapazität der Moos- statistical computing. R Foundation for Statistical und Flechtendecke des Waldes. Veröffentlichung des Computing, Vienna, Austria. Botanischen Instituts der Forstl. Forschungsanstalt Rawls, W.J., Pachepsky, Y.A., Ritchie, J.C., Sobecki, T.M., München. Bloodworth, H., 2003. Effect of soil organic carbon on soil Medina, N., Draper, I., Lara, F., 2011. Biogeography of mosses water retention. Geoderma, 116, 61–76. and allies: Does size matter? Biogeography of microscopic Rice, S.K., Collins, D., Anderson, A.M., 2001. Functional organisms. Is everything small everywhere? Cambridge significance of variation in bryophyte canopy structure. University Press, pp. 209–233. American Journal of Botany, 88, 1568–1576. Morgan, R.P.C., 2005. Soil Erosion and Conservation. 3 edn. Rice, S.K., Gagliardi, T.A., Krasa, R.A., 2018. Canopy Blachwell Publishing, Oxford. structure affects temperature distributions and free Nakatsubo, T., 1994. The effect of growth form on the convection in moss shoot systems. American Journal of evaporation in some subalpine mosses. Ecological Research, Botany, 105, 1499–1511. 9, 245–250. Rice, S.K., Schneider, N., 2004. Cushion size, surface Nebel, M., 2001. Amblystegium serpens (Hedw.) Schimp. In: roughness, and the control of water balance and carbon flux Nebel, M., Philippi, G. (Eds.): Die Moose Baden- in the cushion moss Leucobryum glaucum (Leucobryaceae). Württembergs, Band 2: Bryophytina II, Schistostegales bis American Journal of Botany, 91, 1164–1172. Hypnobryales. Verlag Eugen Ulmer, Stuttgart, pp. 308–309. Richardson, D.H.S., 1981. The Biology of Mosses. Blackwell Nebel, M., Philippi, G., Ahrens, M., Sauer, M., Schäfer- Scientific Publications, Oxford. Verwimp, A., Schoepe, G., 2001. Die Moose Baden- Robinson, S.A., Wasley, J., Popp, M., Lovelock, C.E., 2000. Württembergs, Band 2: Bryophytina II, Schistostegales bis Desiccation tolerance of three moss species from continental Hypnobryales. Verlag Eugen Ulmer, Stuttgart. Antarctica. Australian Journal of Plant Physiology, 27, 379– Niinemets, Ü., Tobias, M., 2014. Scaling light harvesting from 388. moss “leaves” to canopies. In: Hanson, D.T., Rice, S.K. Sayer, E.J., 2006. Using experimental manipulation to assess (Eds): Photosynthesis in Bryophytes and Early Land Plants. the roles of leaf litter in the functioning of forest ecosystems. Advances in Photosynthesis and Respiration (Including Biological Reviews Cambridge Philosophical Society, 81, Bioenergy and Related Processes). Springer, Dordrecht, pp. 1–31. 151–171. Schindelin, J., Arganda-Carreras, I., Frise, E., Kaynig, V., Niinemets, Ü., Tobias, M., 2019. Canopy leaf area index at its Longair, M., Pietzsch, T., Preibisch, S., Rueden, C., higher end: dissection of structural controls from leaf to Saalfeld, S., Schmid, B., Tinevez, J.Y., White, D.J., canopy scales in bryophytes. New Phytologist, 223, 118– Hartenstein, V., Eliceiri, K., Tomancak, P., Cardona, A., 133. 2012. Fiji: an open-source platform for biological-image Novák, V., Hlaváčiková, H., 2019. Applied Soil Hydrology. analysis. Nature Methods, 9, 676–682. Springer, Heidelberg, Berlin. Schneider, C.A., Rasband, W.S., Eliceiri, K.W., 2012. NIH Oishi, Y., 2018. Evaluation of the water-storage capacity of Image to ImageJ: 25 years of image analysis. Nature bryophytes along an altitudinal gradient from temperate Methods, 9, 671–675. forests to the Alpine zone. Forests, 9, 14. Schofield, W.B., 1981. Ecological significance of Price, A.G., Dunham, K., Carleton, T., Band, L., 1997. morphological characters in the moss gametophyte. The Variability of water fluxes through the black spruce (Picea Bryologist, 84, 149–165. mariana) canopy and feather moss (Pleurozium schreberi) Seitz, S., Nebel, M., Goebes, P., Käppeler, K., Schmidt, K., carpet in the boreal forest of Northern Manitoba. Journal of Shi, X., Song, Z., Webber, C.L., Weber, B., Scholten, T., Hydrology, 196, 310–323. 2017. Bryophyte-dominated biological soil crusts mitigate Proctor, M.C.F., 1979a. Structure and eco-physiological soil erosion in an early successional Chinese subtropical adaptation in bryophytes. In: Bryophyte Systematics. forest. Biogeosciences, 14, 5775–5788. Academic Press, London, pp. 479–509. Senf, C., Buras, A., Zang, C.S., Rammig, A., Seidl, R., 2020. Proctor, M.C.F., 1979b. Surface wax on the leaves of some Excess forest mortality is consistently linked to drought mosses. Journal of Bryology, 10, 531–538. across Europe. Nature Communications, 11, 6200. 432 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations Simon, T., 1987. The leaf-area index of three moss species CO2 pulses in biocrusted interspaces in a cool desert (Tortula ruralis, Ceratodon purpureus, and Hypnum ecosystem. Biogeochemistry, 135, 239–249. cupressiforme). In: Pócs, T., Simon, T., Tuba, Z., Podani, J. Voortman, B.R., Bartholomeus, R.P., van Bodegom, P.M., (Eds.): IAB Conference of Bryoecology. Akadémiai Kiadó, Gooren, H., van der Zee, S.E.A.T.M., Witte, J.-P.M., 2014. Budapest-Vácrátót, Hungary, pp. 699–706. Unsaturated hydraulic properties of xerophilous mosses: Slatyer, R.O., 1967. Plant-Water Relationships. Academic towards implementation of moss covered soils in Press, London, New York. hydrological models. Hydrological Processes, 28, 6251– Söderström, L., Hagborg, A., von Konrat, M., Bartholomew- 6264. Began, S., Bell, D., Briscoe, L., Brown, E., Cargill, D.C., Wang, Z., Bader, M.Y., 2018. Associations between shoot-level Costa, D.P., Crandall-Stotler, B.J., Cooper, E.D., Dauphin, water relations and photosynthetic responses to water and G., Engel, J.J., Feldberg, K., Glenny, D., Gradstein, S.R., light in 12 moss species. AoB PLANTS, 10. He, X., Heinrichs, J., Hentschel, J., Ilkiu-Borges, A.L., Weber, B., Berkemeier, T., Ruckteschler, N., Caesar, J., Heintz, Katagiri, T., Konstantinova, N.A., Larraín, J., Long, D.G., H., Ritter, H., Braß, H., Freckleton, R., 2016. Development Nebel, M., Pócs, T., Puche, F., Reiner-Drehwald, E., Renner, and calibration of a novel sensor to quantify the water M.A.M., Sass-Gyarmati, A., Schäfer-Verwimp, A., content of surface soils and biological soil crusts. Methods Moragues, J.G.S., Stotler, R.E., Sukkharak, P., Thiers, B.M., in Ecology and Evolution, 7, 14–22. Uribe, J., Váňa, J., Villarreal, J.C., Wigginton, M., Zhang, Zotz, G., Büde, B., Meyer, A., Zellner, H., Lange, O.L., 1997. L., Zhu, R.-L., 2016. World checklist of hornworts and Water relations and CO2 exchange of tropical bryophytes in liverworts. PhytoKeys, 59, 1–828. a Lower Montane Rain Forest in Panama. Botanica Acta, Soudzilovskaia, N.A., van Bodegom, P.M., Cornelissen, J.H.C., 110, 9–17. 2013. Dominant bryophyte control over high-latitude soil Zotz, G., Schweikert, A., Jetz, W., Westerman, H., 2000. Water temperature fluctuations predicted by heat transfer traits, relations and carbon gain are closely related to cushion size field moisture regime and laws of thermal insulation. in the moss Grimmia pulvinata. New Phytologist, 148, Functional Ecology, 27, 1442–1454. 59–67. Tucker, C.L., McHugh, T.A., Howell, A., Gill, R., Weber, B., Belnap, J., Grote, E., Reed, S.C., 2017. The concurrent use Received 31 March 2021 of novel soil surface microclimate measurements to evaluate Accepted 9 June 2021 SUPPORTING INFORMATION Table S1. Equations of calibration curves for studied soil substrates and moss species. The fit quality is assessed by the root mean square error (RMSE) and the determination coefficient (R ) between measured and modeled water content. Sample Calibration equation a b c d e RMSE R² –07 Angulatensandstein- y = exp(a · x) · b · x + c 0.0043 9.125e 0.264 0.027 0.928 Formation Löwenstein-Formation y = a · x + b (BWP < 1250 mV) 0.00018 0.0322 0.002 0.990 y = a · x + b (BWP > 1250 mV) 0.00215 –2.437 0.016 0.954 –06 Psilonotenton-Formation y = exp(a · x) · b · x + c 0.0038 2.055e 0.221 0.031 0.926 –08 Trossingen-Formation y = exp(a · x) · b · x + c 0.0065 3.218e 0.238 0.010 0.990 Aamblystegium serpens y = (a + b · x) –2.555 0.0045 0.703 0.979 (Lab) Brachythecium rutabulum y = a · x + b 0.0096 –0.401 0.229 0.996 Eurhynchium striatum y = a · x + b 0.0194 –0.617 0.205 0.995 Oxyrrhynchium hians y = a · x + b 0.0127 –0.414 0.200 0.993 Oxyrrhynchium hians (Lab) y = exp(a · x) · b · x + c 0.0008 0.0036 –0.057 0.133 0.998 4 3 2 –11 –08 –05 Plagiomnium undulatum y = a · x + b · x + c · x + d · x + e –3.141e 7.996e –6.066e 0.0198 –0.5402 0.223 0.994 Sonja M. Thielen, Corinna Gall, Martin Ebner, Martin Nebel, Thomas Scholten, Steffen Seitz 434 Water’s path from moss to soil: A multi-methodological study on water absorption and evaporation of soil-moss combinations Figs. S1–S10. Plots of calibration curves for studied soil substrates and moss species. Measured water contents for soil substrates are illustrated in brown and for moss species in green. Table S2. Maximum water storage capacity values (WSCmax) and sample sizes of the studied moss species for immersion and spray tech- –1 nique as gravimetric WSC (g g ), percentage WSC (%) and WSC per unit area (mm), ± standard error of the mean. max max max Moss species Sample WSC WSC WSC WSC WSC WSC max max max max max max –1 –1 size immersion (g g ) spray (g g ) immersion (%) spray (%) immersion (mm) spray (mm) Amblystegium 8 14.097 ± 1.28 10.082 ± 1.25 1409.668 ± 127.82 1008.176 ± 125.09 4.947 ± 0.74 3.144 ± 0.23 serpens (Lab) Brachythecium 8 11.800 ± 0.81 10.049 ± 0.66 1179.965 ± 80.52 1004.919 ± 65.74 3.152 ± 0.31 2.712 ± 0.25 rutabulum Eurhynchium 17 11.223 ± 0.62 9.629 ± 0.40 1122.260 ± 61.55 962.943 ± 39.78 3.342 ± 0.21 2.820 ± 0.18 striatum Oxyrrhynchium 7 9.686 ± 1.41 7.880 ± 0.57 968.598 ± 141.08 787.973 ± 56.90 2.094 ± 0.12 1.945 ± 0.09 hians Oxyrrhynchium 7 9.934 ± 1.24 11.038 ± 1.23 993.381 ± 123.82 1103.796 ± 122.86 2.703 ± 0.32 2.448 ± 0.21 hians (Lab) Plagiomnium 8 7.308 ± 0.80 8.146 ± 0.32 730.792 ± 79.89 814. 613 ± 31.58 1.841 ± 0.29 1.870 ± 0.13 undulatum

Journal

Journal of Hydrology and Hydromechanicsde Gruyter

Published: Dec 1, 2021

Keywords: Biological soil crusts; Bryophytes; Ecohydrology; Moss structure; Moss hydrology; Rainfall interception

There are no references for this article.