The distribution and performance of bryophyte species are known to vary with vertical gradients, due to changes in environmental factors, especially light conditions. However, the morphological and physiological drivers of bryophyte distribution along forest vertical gradients are poorly understood.

Methods: We conducted a comparative analysis of 28 morphological and photosynthetic functional traits in 18 species of mosses and liverworts distributed among three vertical microhabitats (ground, tree trunk, and branch) to analyze trait variance among the microhabitats and bryophyte life–forms in a subtropical cloud forest in Ailao Mountain, Yunnan, southwest China. Principle component analysis (PCA) was used to summarize trait differences among bryophyte species.

Key Results: In contrast to trunk and ground dwellers, branch dwellers tend to reduce light interception (smaller leaf and cell sizes, lower chlorophyll content); protect against damage from intense irradiation (higher ratios of carotenoids to chlorophyll); raise light energy utilization (higher photosynthetic capacity) and cope with lower environmental moisture conditions (pendant life-forms, thicker cell wall). Principal component analysis showed that ecological strategies of bryophytes in response to levels of irradiation were specialized in branch–dwellers, although those in ground and trunk–dwelling species were less distinct.

Conclusions: Environmental filtering shaped the functional traits combination and spatial distribution of bryophytes along the vertical gradients. Bryophyte species from  upper canopy of cloud forests show narrow  variation in functional traits under intense light , whereas species in the lower vertical strata associated with low levels of light intensity exhibited contrasting, but more diverse ecological strategies.

Study area and species

The study area was in a subtropical, evergreen broad–leaved montane cloud forest in the Xujiaba region (23°32′N, 101°01′E) of the Ailao Mountains National Nature Reserve in Yunnan, China, where mean annual precipitation is c. 1859 mm, 86% of which falls from May to October, and mean annual relative humidity (RH) and temperature are 82.6% and 11.6 °C, respectively (Fig. 1). Mean monthly photosynthetic active radiation at 18 m above forest ground layer between September 2014 and August 2017 was higher than at 1.5 m, whereas mean relative humidity was lower (Fig. 1). There is a rich and abundant bryophyte flora in the area, including 176 species belonging to 38 families and 83 genera (Ma et al., 2009), so we selected six dominant species from each of three microhabitats in the forest (ground, tree trunk, and branch) (N = 18) based on species frequency data (Ma et al., 2009; Ma et al., 2011), and classified them according to life–form (after Bates, 1998) (Table 1). The canopy height is 18-25m (Li, 2010) and we mainlycollected samples from dominant trees (Lithocarpus hancei, Castanopsis rufescens, L. xylocarpus, etc.) in the study area and collected bryophyte samples from tree trunks at 0.5-1.5m, and branches at 10-15m above ground, and ground-dwelling bryophytes were collected from forest floor around host trees. Five 20 × 20 m quadrats were randomly placed in a 6-ha plot near Ailao Mountain Station for Subtropical Forest Ecosystem Studies..

Leaf and cell morphological traits

We selected four leaves from five shoots of each species that were mounted and photographed (Olympus IX51 microscope and Olympus DP 80, Tokyo, Japan) at 20–400× magnification. We measured leaf area (LA), leaf length (LL), leaf width at the widest part (LW), length/width ratio (L/W), costa length (CL), and perimeter²/area ratio (P²/A) using ImageJ software (U.S. National Institutes of Health, Bethesda, Maryland, USA), and 4–6 cells from the central area of the lamina were measured to determine cell lumen area (CLA), cell lumen length (CLL), cell lumen width (CLW), and cell wall thickness (CWT); cells per area (CPA) was measured from 15–20 cells by circling a center area (diameter=0.05mm) of leaves.

Photosynthetic light response traits

We collected six replicate samples of shoot material from each species, from March to April 2017, from which photosynthetic light response curves were measured using a portable photosynthesis system (LI–6400, Li–Cor, Inc., Lincoln, NE, USA). After dirty and dead tissue had been removed, 2.5 to 4.5 cm long samples of stem material were cut from the apex or whole plant by removing the rhizoids; the stem material was then submerged and saturated in distilled water for 1 h. Then, excessive superficially adherent water was carefully removed using a paper towel (Zotz et al., 1997), before samples were placed in a whole plant chamber (LI–6400–17) attached to a portable infrared gas analyzer (LI–6400XT) to determine net photosynthetic rates; a full spectrum light source (LI–6400–18) that was placed on top of the chamber provided cold illumination to avoid dramatic changes in temperature. Photosynthetic rate based on bryophyte area and mass is lower than in vascular plants; therefore, we set the flow rate at 300 µmol m⁻² s⁻¹ under ambient environmental conditions (14–20 °C; 50–70% RH). Light response curves were determined using 12 photosynthetically active radiation (PAR) gradients (300, 250, 200, 150, 125, 100, 80, 60, 40, 20, 10, and 0 µmol m⁻² s⁻¹); after measurement of 3-4 PAR gradients, samples were submerged in distilled water for 1 min to maintain hydration. Light response curves were fitted to a modified rectangular hyperbola (Ye, 2007; Ye and Yu, 2008) using equation (1):

A_n=α1_-βI1_-γII-R_area    (1)

where A_n is the net photosynthetic rate derived from irradiance (µmol m⁻² s⁻¹); I is light intensity (µmol m⁻² s⁻¹); R_area is dark respiration rate (µmol m⁻² s⁻¹); and α, β and γ are coefficients. Model fitting of the measured net photosynthetic rate to irradiance was completed using Sigma Plot 11.0 (Systat Software, Inc., San Jose, CA, USA); the software was used to directly fit the coefficients (α, β, γ, and R_d). Subsequently, these coefficients were used to calculate the main photosynthetic parameters, including the light–saturated net photosynthetic rate (A_area, µmol m⁻² s⁻¹), light saturation point (I_s, µmol m⁻² s⁻¹) and light compensation point (I_c, µmol m⁻² s⁻¹).

Shoot mass by area (SMA) was derived as the shoot dry mass divided by the projected area, while net photosynthetic rate by mass (A_mass) and dark respiration by mass (R_mass) were calculated as A_mass = A_area/SMA and R_mass = R_area/SMA, respectively. Projected shoot area was measured using an Epson Perfection V700 Photo flatbed scanner (Seiko Epson Corp., Nagano, Japan) and determined using WinRHIZO Pro 2009b (Regent Instruments, Quebec, Canada). Samples were oven–dried for 48 h at 70 °C after trait measurement to determine dry mass.

Extraction and determination of chlorophyll pigments

Five replicates of approximately 0.3 g of dry leaf tissue per bryophyte species were collected and stored at 4 °C, and then ground under dim light in 95 % ethanol with quartz sand and CaCO₃ powder (0.5g approximately respectively) using a mortar and pestle that had been pretreated at 4 °C. Extracted chlorophyll pigments were measured at 470, 649, and 665 nm absorbance using a spectrophotometer (UV–B 2501, Shimadzu, Japan), and concentration of Chl a, Chl b, carotenoid (Car), and total chlorophyll (Chl) were calculated according to the extinction coefficients of Arnon (Nasrulhaq–Boyce and Mohamed, 1987).

Chlorophyll fluorescence

Chlorophyll a fluorescence signals were measured using a PAM fluorometer (FMS 2, Hansatech, Instruments Ltd., Norfolk, UK) from five replicate leaf samples per species that were saturated in distilled water and dark–adapted for 23 min prior to measurement. Two or three overlapped leaves were selected growing in shoots to fill the measuring area (Φ=0.5cm) (Chen, 2016). Leaf material was exposed to a weak modulated beam to assess initial minimal fluorescence efficiency in the dark–adapted state (F₀), and then a saturation pulse of c. 5500 µmol m⁻² s⁻¹ for 0.7 s was applied to assess maximal photochemical efficiency when PSII reaction centers were closed (F_m). Minimal and maximal Chl fluorescence efficiency and the steady–state Chl fluorescence efficiency in the light–adapted state ((F₀′, F_m′ and F_s) were measured using actinic illumination (c. 110 µmol m⁻² s⁻¹) and saturating illumination, respectively. Maximal photochemical efficiency of PSII in the dark-adapted state (F_v/F_m), where F_v is variable fluorescence yield, was calculated as F_v = F_m−F₀; actual photochemical efficiency of PSII in the light–adapted state (φPSII) was calculated as φPSII = (F_m′−Fs)/F_m′; photochemical quenching (qP) was calculated as qP = (F_m′−F_s)/(F_m′−F₀′); and non–photochemical quenching (NPQ) was derived from NPQ = (F_m−F_m′)/F_m′.