Introduction

Bryophytes dominate the carbon and nitrogen cycling of many poorly drained terrestrial ecosystems (Turetsky 2003; Vitt et al. 2001), particularly in the boreal region, but on a global scale as well (Longton 1992; O’Neill 2000). Such ecosystems—peatlands and to a lesser degree forested wetlands—cover at least 2.5 × 108 ha globally (Gorham 1991; Lugo et al. 1990; Matthews and Fung 1987), store approximately 455 Pg C, mostly in peat, and sequester nearly a third of global soil C (Gorham 1991). Much of the boreal forest is poorly drained, with poorly drained areas exhibiting high rates of primary production (Camill et al. 2001; Vitt et al. 2001), slow decomposition rates and bryophyte-dominated successional pathways (O’Neill 2000; Turetsky 2003). Bryophytes are increasingly recognized to play important roles in the biogeochemical cycling of well-drained boreal forests as well (DeLuca et al. 2002; O’Connell et al. 2003a) and influence the vegetation-atmosphere exchange in many ecosystems (Lafleur and Rouse 1988; Shimoyama et al. 2004; Williams and Flanagan 1996).

Bryophytes are less well studied than vascular plants, and few studies have quantified the bryophyte leaf area index (LAI), the amount of leaf area per unit ground area. These few studies have generally measured the leaf area of individual samples (Proctor 2000; Rice and Schneider 2004; Williams and Flanagan 1998; Zotz et al. 2000), although DeLucia et al. (2003) estimated stand-level bryophyte LAI in a New Zealand forest. In most vascular plant systems, LAI strongly influences energy, water and carbon dioxide exchange between terrestrial ecosystems and the atmosphere and is tightly coupled with photosynthesis, litterfall, microclimate and productivity (Gower et al. 1999). It is an important parameter in terrestrial biogeochemical models and remote sensing (Chen et al. 1997), with many models requiring it as an input, and allows researchers to scale fluxes from the leaf to ecosystem level (Campbell and Norman 1998). Bryophytes differ in many respects from vascular plants—they are generally astomatal, poikilohydric and at least partially ectohydric (carry external water)—but the bryophyte photosynthetic apparatus itself is similar to that of a C3 vascular plant (Green and Lange 1994; Williams and Flanagan 1998). Bryophyte LAI may be much more tightly coupled with carbon than water exchange (Proctor 2000), unlike in most vascular plants, but knowledge of the bryophyte leaf area would be valuable for studies at the plant, ecosystem and landscape levels.

Many methods have been used to estimate the leaf area of large vascular plants (Bonan 1993; Chen et al. 1997); most bryophyte studies of leaf area have used contact probes (Rice and Schneider 2004) or light attenuation measurements (DeLucia et al. 2003). Leaf area may also be calculated from leaf mass using specific leaf area (SLA), the ratio of the fresh foliage surface area to the unit dry foliage mass. In vascular plants SLA typically varies with tree species and environmental conditions, decreases with foliage age and higher canopy position, and is correlated with the potential growth rate (Lambers and Poorter 1992) and leaf longevity (Gower et al. 1993; Reich et al. 1992, 1997). Many studies of bryophyte biomass have been performed, and knowledge of bryophyte SLA could provide bryophyte leaf area estimates for many areas. We know of no published measurements of bryophyte SLA, however.

The goals of this study were to quantify bryophyte spatial coverage, live biomass, SLA and thus LAI for a group of boreal forest stands. This was conducted on a species-specific level as bryophyte diversity and morphology are highly variable, as in vascular plants (Shaw and Goffinet 2000). Because structural parameters such as SLA can change drastically with water status (Tuba et al. 1996), bryophyte water content was also sampled and reported.

Methods

Site descriptions

The study was conducted in four different-aged (16, 41, 75 and 155 years in 2005) boreal forests west of Thompson, Manitoba, Canada; the oldest stand (at 55°53′N, 98°20′W) is the BOREAS NSA tower site (Halliwell and Apps 1997). Separate well and poorly drained stands (about 50 m × 50 m) were located in each forest. The overstory in the two oldest stands was primarily black spruce; in the younger two stands, trembling aspen (Populus tremuloides Michx) and jack pine (Pinus banksiana Lamb.) were also present (Bond-Lamberty et al. 2002b). In well-drained stands of central Manitoba, the black spruce canopy closure, at 50–60 years, is associated with drastic thinning of the understory and growth of thick feather mosses; in the poorly drained stands, the canopy remains open, with black spruce, Labrador tea (Ledium groenlandicum Oeder), and Sphagnum spp. dominating biomass and net primary production (Bond-Lamberty et al. 2004; Gower et al. 1997; Wang et al. 2003). Mean annual temperatures were −1 to −2°C. Select stand characteristics are summarized in Table 1.

Table 1 Stand characteristics, by soil drainage and stand age (years), with standard errors in parentheses
Full size table

Field sampling protocol

In each stand, the spatial coverage of bryophyte species was assessed by randomly placing a 0.25 m2 sampling ring on the ground (n = 100) and visually estimating the percent cover of each species in the ring. Approximately 1% of a 50 m × 50 m area, i.e., 25 m2, was sampled in this way. We did not attempt to enumerate every single species present, as some bryophyte species require microscopic analysis to distinguish, and others are extremely rare. Instead, we focused on the three to six species in each stand that dominated (95%+) spatial coverage, and presumably biomass, leaf area and net primary production.

Weft-forming and pleurocarpous feather mosses categorized were Schreber’s big red stem moss [Pleurozium schreberi (Brid.) Mitt.], knight’s plume moss [Ptilium crista-castrensis (Hedw.) De Not.] and splendid feather moss [Hylocomium splendens (Hedw.) Schimp. in B.S.G.]. In the lowlands, Sphagnum species enumerated included S. fuscum (Schimp.) Klinggr., S. riparium Ångstr. and S. warnstorfii Russ., S. angustifolium (C. Jens. ex Russ.) C. Jens. in Tolf and S. capillifolium (Ehrh.) Hedw. Also common in both well and poorly drained stands was glow moss, Aulacomnium palustre (Hedw.) Schwaegr. This bryophyte has a similar growth form to, and frequently grows intermixed with, the brown moss Tomentypnum nitens (Hedw.) Loeske (Johnson et al. 1995); we did not attempt to separate these two species. An ‘other’ category was used for infrequently occurring bryophytes, in these stands most often Polytrichum spp. The names and taxonomy given here follow the USDA plants database (http://www.plants.usda.gov).

Five random samples, each 10 cm × 10 cm and generally approximately 10 cm depth, were taken of each major (>5% coverage) bryophyte species at the three oldest stands. Sampling occurred over a 4-day period in August 2005 during which there was no rain; meteorological records from Thompson, MB (40 km east) indicated 21 mm of rain in the previous 7 days and 135 mm in the previous 30 days (D. Bronson, University of Wisconsin, unpublished data). Samples were stored in airtight plastic bags, chilled and processed in the lab within 24 h.

Laboratory protocol

Samples were sorted, weighed and scanned in the laboratory. Each sample was separated into live and dead tissue based on color, depth and decay. Treatment of samples was species-specific:

  • H . splendens: This species was separated into live and dead tissue, and the live tissue separated into 1st- and 2nd-year growth; Hylocomium is the only bryophyte sampled with such a determinate growth form (Fig. 1). Leaves and secondary costae were separated from the main midrib.

  • A. palustre/T. nitens, P. schreberi, P. crista-castrensis: The top approximately 1.5 cm was treated as live, following Williams and Flanagan (1998) and field observations (Bond-Lamberty, pers. obs). For P. schreberi, H. splendens and A. palustre, the main midrib was separated from secondary costae and scanned and weighed separately.

  • S. fuscum: The very compact growth form of this species allows little light to penetrate below the capitulum, and thus the top 0.5 cm was treated as the live (photosynthesizing) plant. The branches of S. fuscum are densely interleaved in fascicles and were separated before scanning (Fig. 1).

  • S. warnstorfii, S. riparium, S. capillifolium: These species are less densely arranged than S. fuscum, and the top approximately 2.5 cm were treated as live. Branches were separated before scanning.

Bryophyte samples were scanned at 600 DPI using a backlit flatbed scanner (Canon LiDE, Canon USA, Lake Success, NY). The resulting TIFF image files were converted from grayscale to black and white using a threshold luminosity value of 175 (on a 0–255 scale). This threshold value yielded pixel counts virtually identical to a subset of images in which pixels were counted by hand under both low-shadow (y = 0.9642x, R 2 = 0.997, n = 10) and high-shadow (y = 0.7132x, R 2 = 0.991, n = 10) conditions (Bond-Lamberty, unpublished data). Two separate regressions were used because some samples exhibited pronounced shadowing, caused by the small bryophyte leaves not contacting the scanner bed cover, but instead projecting their shadows upon it. Images were subjectively categorized as “low-shadow” and “high-shadow,” and a hand count was performed if there was any doubt. Approximately 13% of the samples (n = 19) were hand-counted, with pixel counts on average 1% (±SD of 11%) below the regression-based value. Projected leaf area (PLA) was computed by dividing the final image pixel count (based on the appropriate regression, or the hand count) by the scan resolution in pixels per unit area.

Fig. 1
figure 1

Microscope images of bryophyte samples. Top row, left to right: Pleurozium schreberi at 7.5×; P. schreberi at 30×; Hylocomium splendens’ determinate growth form at 7.5×. Bottom row, left to right: Sphagnum fuscum at 7.5×; separated branches of S. fuscum at 7.5×; detail of S. fuscum branch at 64×. Scale ruler (orange lines) is in mm

Full size image

Leaf area index (LAI) was defined as hemisurface leaf area (HSLA), or one-half total leaf area per unit ground area. The conversion of PLA to HSLA is dependent on the cross-sectional leaf geometry (Bond-Lamberty et al. 2003; Chen et al. 1997). HSLA for flat-leaved species (the three feather mosses H. splendens, P. schreberi and P. crista-castrensis, and A. palustre) was assumed to be equal to PLA (i.e., HSLA:PLA = 1). A dissecting microscope was used to examine Sphagnum branch cross-sections (Fig. 1). Based on these images and the known morphology of Sphagnum (Schofield 2001), a semicircular geometry was used for all Sphagnum species (HSLA:PLA = π/2).

All samples were dried to a constant mass in a forced-air oven at 70°C, weighed and the water content computed on a dry-mass basis (fresh mass minus dry mass, divided by dry mass). Specific leaf area (SLA) was computed as HSLA (cm2) divided by leaf mass (g). Area-normalized biomass (g C m−2) was computed using the spatial survey and biomass data; for species with non-photosynthesizing midribs (P. schreberi, A. palustre and H. splendens), foliage and total biomass (B f and B t) were calculated separately. The leaf area index (LAI), defined as HSLA per unit ground area, was computed by multiplying foliage biomass by SLA for each species. Site-specific SLA values were used when possible; if a particular species was present at a site, but its SLA was not sampled at that site, the mean from other sites was used.

Statistical analyses were conducted using SAS version 8.1 software (SAS Institute Inc. 2001). Post-hoc tests for species and site differences were performed using a Tukey-Kramer protected test in PROC MIXED with a base α level of 0.05; unless otherwise noted, all calculations and statistical analyses used bryophyte sample as the experimental unit (n = 100) and a significance level of α = 0.05.

Results

Coverage, structure and water content

P. schreberi dominated bryophyte ground cover in the two oldest well-drained stands (69–84%) and was an important component in the two oldest poorly drained stands (Table 2). P. schreberi spatial coverage was significantly related to overstory leaf area (y = −3.2 + 12.4x, R 2 = 0.93; P = 0.0001; data not shown). A. palustre was the dominant bryophyte in the younger stands; Sphagnum species, particularly S. fuscum, were important in the poorly drained stands (Table 2). Bryophyte diversity was higher in the poorly drained stands than in the well drained ones, and monotonically increased with stand age in the poorly drained stands.

Table 2 Spatial coverage of brophyte species in study sites (%), by soil drainage and stand age (years)
Full size table

Three of the bryophyte species sampled (A. palustre, H. splendens and P. schreberi ) had relatively tough, non-photosynthesizing main midribs, and leaves and midrib masses were determined separately. Leaf mass as a percentage of total sample mass was 37 ± 5% (mean ± standard deviation, n = 10) for A. palustre, 54 ± 12% (n = 8) for H. splendens and 55 ± 9% (n = 18) for P. schreberi. These percentage did not differ significantly between stands (F 3,12 = 1.81; P = 0.199; two stands for A. palustre and H. splendens, four stands for P. schreberi). Leaves of A. palustre comprised a significantly smaller fraction of plant mass than did those of H. splendens or P. schreberi (F 1,12 = 17.31; P = 0.001).

Water content in situ—defined as bryophyte fresh mass minus dry mass, divided by dry mass—did not differ between the live moss at the surface and dead tissue underneath (F 1,151 = 0.01; P = 0.925). Water content did differ significantly between species, even after taking into account site-to-site variability (F 6,151 = 100.48; P < 0.001); from highest to lowest water content, the species’ order in this regard was S. riparium (2,066% ± 520%, n = 23) > S. capillifolium > S. warnstorfii > S. fuscum > A. palustre > P. crista-castrensis > P. schreberi > H. splendens (400 ± 92%, n = 73). The well-drained 41-year-old stand had drier bryophytes than the others (T 151 = −2.09; P = 0.039); otherwise, there was no difference between stands in this regard.

Biomass, specific leaf area and leaf area index

Total live bryophyte biomass in the well-drained stands ranged from 46.6 g m−2 at the 41-year-old stand to 229.9 g m−2 at the 155-year-old one (Fig. 2). Biomass in the poorly drained stand was generally higher, from 102.1 g m−2 in the youngest stand to 228.0 g m−2 at the oldest. The dominance of A. palustre, H. splendens and P. schreberi meant that in most stands total live biomass was equally comprised of leaf and non-leaf tissue.

Fig. 2
figure 2

Bryophyte biomass by stand age, for a well-drained and b poorly drained stands. “Nonleaf” refers to main midrib of P. schreberi, H. splendens and A. palustre. Error bars show standard deviation of total biomass (n = 100 samples, each 0.25 m2)

Full size image

Specific leaf area of samples varied between 135.1 and 472.9 cm2 g−1; these lowest and highest values were for A. palustre and S. capillifolium, respectively (both at the 41-year-old poorly drained stand; Table 3). SLA was strongly affected by bryophyte species across the entire data set (F 7,61 = 26.44; P < 0.0001), but not by stand (F 5,61 = 1.10; P = 0.374); the species-stand interaction was not significant. For individual bryophyte species, the only significant difference in SLA among stands was for A. palustre (135.1 versus 206.3 and 212.8 cm2 g−1; F 2,9 = 13.17; P = 0.004; Table 3). In general, there was no significant difference between the SLA of non-Sphagnum mosses.

Table 3 Specific leaf area (cm2 g− 1, hemisurface basis) for bryophyte species, by soil drainage and stand age (years)
Full size table

The midrib of P. schreberi had significantly lower SLA (difference of 91.5 cm2 g−1; F 1,17 = 23.72; P = 0.0001) than did the leaves, as did the midrib of A. palustre (difference of 61.4 cm2 g−1; F 1,11 = 24.56; P = 0.0004). The midrib of H. splendens also had lower SLA than its leaves (by 39.1 cm2 g−1), but the difference was not significant (F 1,7 = 2.67; P = 0.146). Sphagnum species had much more variable, and significantly (F 2,51 = 24.29; P < 0.0001) higher, SLA than did the feather mosses (Table 3). Leaf age (current year versus previous year) had no significant effect on SLA of H. splendens (F 1,8 = 0.01; P = 0.920).

Bryophyte leaf area index increased with stand age in both the well and poorly drained stands, peaking at 3.1 and 3.7, respectively (Fig. 3). Bryophyte LAI comprised 28–58% of overstory LAI in the well-drained stands, but in the poorly drained stands bryophyte LAI exceeded overstory LAI (by 336–1,133%) in every stand (Bond-Lamberty et al. 2002b). Bryophyte LAI was dominated by P. schreberi in the well-drained stands and Sphagnum spp. in the poorly drained stands. Total shoot area index (SAI, i.e., leaf + nonleaf area, hemisurface basis) was 2.0–5.0 in the well-drained stands and 1.6–4.9 in the poorly drained stands (data not shown; SAI increased with stand age in both cases).

Fig. 3
figure 3

Bryophyte leaf area index (LAI, m2 m−2), hemisurface basis, by stand age, for a well-drained and b poorly drained stands. Line graphs show overstory LAI for each stand (Bond-Lamberty et al. 2002b). Error bars show standard deviation (n = 100 samples, each 0.25 m2)

Full size image

Discussion

Bryophytes have a large impact on the carbon, nitrogen and water cycling of northern ecosystems (O’Neill 2000; Turetsky 2003), with different functional groups (i.e., feather mosses and Sphagnum) having significantly different effects (Bisbee et al. 2001). The spatial coverage and moss biomass numbers reported in this study are consistent with previous studies in these sites (Gower et al. 1997; Wang et al. 2003) and other boreal forests (Longton 1992; O’Connell et al. 2003b; O’Neill 2000). Bisbee et al. (2001) reported Sphagnum percent cover was inversely correlated to overstory LAI for a boreal Saskatchewan forest, much as the current study found Pleurozium cover positively correlated to overstory LAI. Although this study sampled only a small area (2,500 m2) in each of a few boreal forest sites, the stands studied here have been shown to be representative of a larger group of forest stands sampled throughout north-central Manitoba (Bond-Lamberty et al. 2004).

Bryophyte canopy structure and leaf area

Bryophyte canopy structure strongly affects water uptake, storage and loss and varies considerably among species (Rice et al. 2001). A key structural metric is SLA; the SLA values reported here spanned a wide range (135–473 g cm−2), especially within the Sphagnum samples (Table 3). Light and water exert strong influences on Sphagnum canopy structure (Hayward and Clymo 1983), and variation with a species can be caused by, e.g,. ultraviolet radiation in S. fuscum (Gehrke 1998). The values in Table 3 may represent the low end of bryophyte SLA, particularly as we did not measure at the scale of individual single-cell branch leaves. Rice and Schneider (2004) used a contact probe to measure SLA of approximately 1,700 cm2 g−1 for the cushion moss Leucobryum glaucum (S. Rice, personal communication), an order of magnitude larger than the values reported here.

The large contribution of bryophyte LAI to overall stand LAI, and importance of bryophyte to boreal carbon (Harden et al. 1997) and water (Price et al. 1997) fluxes, emphasize the need to include water and carbon fluxes of bryophytes in many forest ecosystem analyses. However, few direct measurements of bryophyte leaf area have been made. DeLucia et al. (2003) estimated a total LAI of 6 for bryophytes in a mature New Zealand forest, based on light attenuation through the bryophyte canopy. The LAI values reported here are expressed as hemisurface leaf area, i.e., one-half of the total leaf area, and thus our oldest stands’ values (3.1–3.7) are similar to those measured by DeLucia et al. (2003). We are unaware of other previous published estimates of stand-level bryophyte leaf area, although values for individual samples have been reported. Proctor (2000) and Proctor and Tuba (2002) estimated LAI values of 6 for Tortula (Syntrichia) intermedia, 18 for Mnium hornum and 20–25 for Scleropodium (Pleurscleropodium) purum. Williams and Flanagan (1998) measured leaf area of P. schreberi from a black spruce site in Saskatchewan. They did not report these data, but imply that a total shoot area index of 2.4 was typical (Williams and Flanagan 1998), overlapping our data from the 155-year-old stand, samples from which had (hemisurface) LAI values ranging from 0.5 to 5.3 and SAI from 0.6 to 6.3.

The high LAI values noted above are characteristic of bryophytes (Proctor 2000), the result of a different set of ecophysiological tradeoffs than those faced by large vascular plants. High LAI certainly maximizes light interception for photosynthesis (Whitehead and Gower 2001); it may also mitigate the slow (by a factor of 10−4 as compared to air) molecular diffusion of CO2 through these plants’ internally and externally carried water volumes (Proctor 2000). Evaporation from bryophyte leaves is controlled by diffusion through boundary layers adjacent to plant surfaces, as in vascular plants (Campbell and Norman 1998). However, (1) the general absence of stomata mean that the boundary layer is the major impediment to water loss (Proctor 1980) and (2) bryophyte cushion size and surface roughness control boundary layer thickness—and thus water relations and carbon gain (Zotz et al. 2000)—as the small bryophyte leaves lie entirely within the cushion’s laminar boundary layer at low to moderate wind speeds (Rice and Schneider 2004). Thus, in most conditions the bryophyte colony surface functions as a single “leaf” lacking an outer epidermis (Proctor 2000), enabling the bryophyte to maintain a large leaf area, while paying a smaller penalty in water loss than would most vascular plants. Their impact on ecosystem water cycling is nonetheless significant: boreal bryophytes were shown to intercept 23% of water throughfall (compared to 15–60% for the overstory canopy), with most of that subsequently evaporating, in a Picea mariana canopy near the sites described here (Price et al. 1997).

Sources of error and methodology assessment

Several potential sources of error can be identified in our measurements and estimates of stand-level bryophyte LAI. First, the scale of measurement is a significant factor with bryophytes, whose functional units span three orders of magnitude (0.1 to 10+ mm) (Proctor and Tuba 2002; Rice et al. 2001). It is important to note that our measurements were not at the scale of the actual branch leaves in some of these species (Fig. 1), affecting this method’s accuracy. For example, the actual branch leaves of Sphagnum are one cell thick and overlap (Schofield 2001), and the actual hemisurface area of such cells can be expected to be larger (and thus the SLA and LAI higher) than the values given here. Higher PLA:HSLA conversion factors than the ones used here, which assumed smooth surface geometries, could be used to convert projected to hemisurface leaf area at this small scale.

A second problem is one of sampling design, not methodology: as noted above, the results given for A. palustre actually include a mix of this species and T. nitens. These two mosses have some similarities (and in this study there was no difference in non-Sphagnum SLA), but belong to different taxonomic orders. Given the ubiquity of these species in boreal fens and swamps (Johnson et al. 1995), and the great variability in SLA measured here (Table 3), species-specific data could be useful.

Thus, there are reasonable questions about the accuracy of our results. The precision of this method, however, is high: biomass, projected area and cross-sectional morphology can all be determined with relatively little error. In addition, once SLA values are determined, estimation of bryophyte LAI simply requires an accurate biomass estimation—a simple and routinely made measurement, and one that integrates over the entire stand area. In contrast, techniques such as contact probes (Rice and Schneider 2004) and light extinction measurements (DeLucia et al. 2003) are single-point measurements that require more time and money to perform well. Direct comparisons of bryophyte leaf area determined by these different methods would be a useful next step in accurately estimating stand-level bryophyte LAI.

Finally, the ease of LAI estimation using this method may in turn promote new research possibilities. For instance, moss flux measurements are typically expressed on a ground area basis, unlike the leaf-level measurements performed for vascular plants, but knowledge of moss LAI would allow for more direct comparisons to be made between the leaf gas exchange rates of these groups. Currently, most studies of evapotranspiration do not consider bryophyte leaf area, even when measuring moss-dominated systems (Shimoyama et al. 2004). Mosses and other bryophytes play significant roles in the biogeochemical cycling of boreal systems, and a simple way to estimate their leaf area would improve our understanding of these processes.