Bryophyte physiological responses to, and recovery from, long-term nitrogen deposition and phosphorus fertilisation in acidic grassland


Author for correspondence:
María Arróniz-Crespo
Tel:+44 114 222 0082
Fax: +44 114 222 0002


  • • Atmospheric nitrogen deposition can cause major declines in bryophyte abundance yet the physiological basis for such declines is not fully understood. Bryophyte physiological responses may also be sensitive bioindicators of both the impacts of, and recovery from, N deposition.
  • • Here, responses of tissue nutrients (nitrogen (N), phosphorus (P) and potassium (K): NPK), N and P metabolism enzymes (nitrate reductase and phosphomonoesterase), photosynthetic pigments, chlorophyll fluorescence, sclerophylly and percentage cover of two common bryophytes (Pseudoscleropodium purum and Rhytidiadelphus squarrosus) to long-term (11 yr) enhanced N deposition (+3.5 and +14 g N m−2 yr−1) are reported in factorial combination with P addition. Recovery of responses 22 months after treatment cessation were also assessed.
  • • Enhanced N deposition caused up to 90% loss of bryophyte cover but no recovery was observed. Phosphomonoesterase activity and tissue N : P ratios increased up to threefold in response to N loading and showed clear recovery, particularly in P. purum. Smaller responses and recovery were also seen in all chlorophyll fluorescence measurements and altered photosynthetic pigment composition.
  • • The P limitation of growth appears to be a key mechanism driving bryophyte loss along with damage to photosystem II. Physiological measurements are more sensitive than measurements of abundance as bioindicators of N deposition impact and of recovery in particular.


Anthropogenic nitrogen (N) deposition poses a global threat to ecosystems and their floristic diversity (Sala et al., 2000; Phoenix et al., 2006). While much research to date has focused on the responses of higher plants, bryophytes have received far less attention despite these being major components of many ecosystems and known to be particularly sensitive to acid deposition (Bates, 2002). Furthermore, while major declines in bryophyte abundance resulting from increased N deposition have been reported (Lee & Caporn, 1998; Gunnarsson & Rydin, 2000; Bergamini & Pauli, 2001; Gordon et al., 2001; Pearce & van der Wal, 2002; Carroll et al., 2003; Pearce et al., 2003), elucidation of physiological responses driving such bryophyte losses have tended to focus on one or a few variables, often tissue N and the resulting N saturation measured through nitrate reductase activity (Gordon et al., 2001; Pearce & van der Wal, 2002; Pearce et al., 2003). While such studies have been invaluable in providing the first insights into likely causes of bryophyte decline, we currently lack a much broader analysis of physiological responses (such as responses to increased P limitation, photosystem damage, stress and shading) which would allow us to determine which variables are the most important drivers of bryophyte loss and understand their relative importance compared with those measured previously.

The importance of this fuller elucidation of bryophyte responses has further value since the sensitivity of bryophytes to N deposition means they are likely to be particularly useful sensitive bioindicators of the impacts of N deposition in ecosystem surveys and may provide earlier warning of the onset of N deposition impacts in comparison with higher plants. Physiological responses may also be particularly useful here since they may be more responsive and readily detectable, especially compared with measurements of growth or abundance, which can be harder to quantify with high precision in bryophytes. Furthermore, given that some previously heavily polluted areas (such as parts of the UK: NEGTAP, 2001; Fowler et al., 2004) are now experiencing declines in N deposition rates, as legislation and emission abatement strategies begin to take effect, bryophytes may also be valuable as sensitive and early indicators of ecosystem recovery from N deposition.

A key cause of bryophyte sensitivity to N deposition is the absence of a well-developed cuticle and hence their capacity to absorb nutrients and atmospheric pollutants over their entire surface area (Bates, 2002). This can result in increased tissue N and N : P ratios (Bragazza et al., 2004; Pitcairn et al., 2006) which may cause potassium (K) leakage where loss of membrane integrity results from N toxicity (Pearce et al., 2003). Changes in soluble metabolites may also result from increased N loads (Baxter et al., 1992; Paulissen et al., 2005; Koranda et al., 2007) while N saturation of the bryophyte can lead to loss of inducible nitrate reductase activity (Woodin & Lee, 1987; Pearce et al., 2003).

Surprisingly, few studies have assessed the effect of N deposition on photosynthesis or photosynthetic machinery, despite the likelihood that this is a sensitive target given that most leaf N is invested here (Lambers et al., 1998). Nitrogen pollution may also indirectly affect bryophytes through increased shading where higher plant cover increases (Bergamini & Peintinger, 2002; van der Wal et al., 2005) and may also exacerbate phosphorus (P) limitation of plant growth, the alleviation of which (though P additions) can counteract N deposition effects in higher plants (Phoenix et al., 2003) and Sphagnum species (Limpens et al., 2004).

Despite this knowledge, no broad analysis of multiple physiological responses on more than one species within the same ecosystem exists to assess the relative importance of different physiological responses in driving bryophyte loss and which may represent the best indicators of N deposition impacts and recovery.

With these concerns in mind, here we report a wide-ranging analysis of multiple physiological responses of two common terrestrial European bryophytes of contrasting growth form to long-term (11 yr) N deposition and recovery from N deposition in field plots (22 months ambient deposition following cessation of treatments). We also assess how these responses are effected by P fertilization since N deposition can enhance P limitation – a limitation that bryophytes may be particular susceptible to given their lack of roots and reliance on atmospheric deposition for nutrients.

Since any broad analysis cannot include measures of all possible parameters, our measurements were chosen to include those most commonly made in past studies and new measures that we predicted to be both responsive to N deposition and be likely drivers of bryophyte loss. Specifically, we assessed the nutritional status and requirements of the bryophytes through tissue chemistry (NPK) and activities of inducible nitrate reductase and phosphomonoesterase. We included tissue K since membrane damage can lead to K leakage. Furthermore, we undertook in vivo chlorophyll fluorescence measurements to evaluate the physiological health of the plants and impacts on photosynthetic machinery; we also assessed changes in the composition of photosystem II (through photosynthetic pigment analysis) and responses to the light environment where the mosses were growing (using de-epoxidation of the xanthophyll cycle pool (DES) and nonphotochemical quenching (NPQ)). We measured the ‘sclerophylly index’ to determine whether N deposition altered the proportion of nonphotosynthetic tissue. These physiological studies were coupled with measurements of the abundance of the bryophyte species under a range of different N deposition, P addition and recovery treatments in long-term field plots.

The study was undertaken in acidic Festuca–Agrostis–Galium grassland classified as U4e in the National Vegetation Classification (Rodwell, 1992), and located in the Derbyshire Dales National Nature Reserve, UK. This is the most widespread type of acid grassland in the UK and is known to be sensitive to N deposition; this type of grassland is in decline throughout Europe as a result of agricultural improvement and N deposition impacts (Preston et al., 2002; Carroll et al., 2003; Stevens et al., 2004). The Derbyshire Dales area has historically received among the highest rates of N deposition in the UK, with a cumulative total of up to 3000 kg N ha−1 over the last century (Fowler et al., 2004). The field site still experiences a high ambient deposition of c. 2.5 g N m−2 yr−1.

By analysing multiple physiological responses in the same two species at the same time we are able to (1) determine the relative sensitivity of different physiological variables to N deposition and recovery, (2) relate these to changes in abundance and hence begin to propose physiological mechanisms behind bryophyte species change, and (3) determine which physiological responses may provide the most sensitive indicators of both impacts of N deposition and the onset of recovery.

Materials and Methods

Study species

Pseudoscleropodium purum (Hedw.) Fleisch. grows mostly in single species mats in contact with the litter layer; it is a slow-growing desiccation-tolerant species with a weft life-form, poorly developed rhizoids and an opportunist strategy for absorbing nutrients (Rincón, 1988; Bates, 2000). Rhytidiadelphus squarrosus (Hedw.) Warnst. grows upwards, usually in single species mats; it is a relatively fast-growing species with a turf life-form with well-developed rhizoids and a high cation exchange capacity for nutrient uptake. Both species are common mosses of European grasslands (Bates, 2000) where they can occur as intimate mixtures with the other vegetation.

Field plots and simulated N deposition

The acid grassland is sited on Wardlow Hay Cop (300 m a.s.l.), a small conical hill in the Derbyshire Dales, UK (53°15°44N, 1°44′02W). Further details of the grassland flora can be found in (Morecroft et al., 1994).

In July 1995, 3 × 3 m plots were established to which N and P treatments were applied following a randomized block design (3 N × 2 P × 3 blocks). Treatments were applied as NH4NO3 at 0 (distilled water only), 3.5 and 14 g N m−2 yr−1 with and without P (as NaH2PO4·H2O) at 3.5 g P m−2 yr−1. Ambient N deposition at the site is c. 2.5 g N m−2 yr−1 with a past maximum of c. 3.5 g N m−2 yr−1 in the late 1980s (INDITE, 1994). Treatment applications were made quarterly in the first year and at monthly intervals thereafter (2 l water per plot per application) using backpack sprayers (Bastion 15; Application Techniques Ltd, Hassocks, Sussex, UK).

To determine the recovery of these grasslands from N deposition, in January 2005, each plot was split in half and treatments ceased in one half (December 2004 therefore being the last month treatment was applied). These untreated half plots are from now on referred to as ‘recovery plots’. Results represent responses of bryophytes to either 11 yr plus 4 months continuous treatment or 9 yr plus 6 months treatment followed by 22 months’ recovery.

Plant samples were collected over two consecutive weeks in October 2006 (a month when both species show good physiological vitality and avoiding the less active periods in the drier summer months). Plant material was collected into plastic bags and immediately transported to the laboratory where the apical green part of each shoot was removed for physiological measurements. Apices were stored in the dark at 4°C in Petri dishes containing a moist cotton-wool base and all physiological variables were measured within 24 h after collection with the exception of tissue nutrient content. For this, samples were directly dried at 80°C for 48 h and stored until further analysis (Martínez-Abaigar et al., 2002).

Bryophytes abundance and physiological measurements

Bryophyte cover  Percentage cover estimates of the two bryophyte species were undertaken using a 50 × 50 cm quadrat placed six times on a transect down the centre of each plot.

Sclerophylly index (SI)  SI was calculated as the ratio between the dry weight and surface area of the fresh prostrate apex (measured using a LI-3000 area meter; LI-COR, Lincoln, NE, USA). In bryophytes this is used as an indicator of the proportion of nonphotosynthetic tissue (Martínez-Abaigar et al., 1994).

Shoot nutrient content  The N, P and K in oven-dried apical shoots were determined following Kjeldahl digestion (Allen, 1989). The N and P content were determined colorimetrically by flow injection analyser (FIAflow2; Burkard Scientific, Uxbridge, UK) and K by a model 2100 atomic absorption spectrophotometer (Perkin-Elmer, Wilton, CT, USA).

Phosphomonoesterase (PME) and nitrate reductase (NR) enzyme activities  The PME activity was determined on two to three fresh shoot apices (c. 6 mg DW equivalent) by measuring the release of p-nitrophenol (p-NP) from p-nitrophenyl phosphate (p-NPP; Sigma substrate 104; Sigma-Aldrich, Poole, UK), as described by Phoenix et al. (2004). The buffer was adjusted to pH 5 as this pH was previously determined to be the optimum pH for PME activity in these bryophytes.

Induced NR activity was determined following Woodin & Lee (1987) and Pearce et al. (2003) on 30–40 mg dry weight equivalent fresh apical samples. After the induction period (6 h dark storage in 5 ml 3 mm KNO3), samples were vacuum infiltrated with 5 ml buffer containing 50 mm KH2PO4, 100 mm KNO3, 100 mm potassium acetate and 1.5% v : v propanol-1-ol, and placed in a dark water bath at 30°C for 30 min. After cooling, nitrite production was qualified by removing 1 ml solution and adding this to 1 ml of 1% (w : v) sulphanilic acid in 1.5 m HCl and 1 ml of 0.02% (w : v) n-(1-napthyl ethylene diamine-HCl) and storing for 40 min in the dark before measurement of absorbance at 540 nm (A540) in a spectrophotometer (Cecil CE1020, Cambridge, UK).

For both assays, dry weight of shoot apices were determined by oven drying tissue samples at the end of the assay.

Pigment composition  The tissue pigment concentrations of chlorophyll (chl), including Chl a/b ratio and carotenoids, were measured to show changes in the composition of PSII and as indicators of stress, and xanthophyll cycle carotenoids as indicators of the light environment experienced by the bryophytes. For these analyses, shoot apices (c. 3 cm2 of fresh tissue per sample) were frozen in liquid N immediately after collection and stored in the dark at –80°C. Samples were powdered with liquid nitrogen in a mortar and extracted with 2 ml 100% cooled acetone (approx. 4°C). Sodium ascorbate was added in order to negate traces of acid in the acetone (Gogorcena et al., 2001). After extraction for 2 h at 4°C the homogenate was filtered through 0.45 µm GF/C Whatman filters (Whatman, Clifton, NJ, USA). Absorption at 661.6 and 644.8 nm of 1 ml sample was read in a spectrophotometer (Perkin-Elmer λ35BUV/Vis) to calculate the amount of chl a and b on the basis of the Lichtenthaler equation (Lichtenthaler, 1987). The rest of the extract was analysed by high-pressure liquid chromatography (HPLC) for carotenoid concentration following Farber et al. (1997). Carotenoids were detected with a photodiode array detector at 450 nm (ref. 750 nm) and quantified by calibration curves of commercial standards of lutein, zeaxanthin and β-carotene (CaroteNature, Lupsingen, Switzerland). The (neoxanthin + lutein)/β-carotene ratio was calculated to estimate differences in the proportion of the light-harvesting complexes (LHC) to reaction centres (RC) between the treatments as neoxanthin and lutein are both carotenoids mainly present in the LHC antenna and β-carotene in the core antenna complexes of RC (Horton et al., 1996, and references therein). The extent of de-epoxidation of the xanthophyll cycle pool (DES) was used as an estimate of the extent of light exposure (Demmig-Adams & Adams III, 1996) and was estimated from the levels of violaxanthin, antheraxanthin and zeaxanthin calculated as (antheraxanthin + zeaxanthin)/(violaxanthin + antheraxanthin + zeaxanthin).

In vivo chlorophyll fluorescence  The maximum quantum yield of PSII (Fv/Fm) was measured to assess damage to photosystem II (PSII) and as a proxy for stress. The effective quantum yield of PSII during actinic illumination (ΦPSII) was used as an indicator of the efficiency of light utilization in photosynthesis and maximal nonphotochemical quenching (NPQmax) was used as an indicator of the light environment. Chlorophyll fluorescence measurements were undertaken using a MINIPAM portable pulse amplitude modulation fluorometer connected to a 2060-M micro quantum/temperature sensor (Walz, Effeltrich, Germany) following (Schreiber et al., 1995). Minimal and maximal fluorescence (F0 and Fm) were measured in samples dark-adapted for 20 min, using a 600 Hz modulated beam at 0.04 µmol m−2 s−1 photon flux density (PFD), and ‘white’ saturating flashes of 10 000 µmol m−2 s−1 PFD and 0.8 s duration, respectively. White actinic light was then applied with incremental PFD (0, 30, 100, 200, 400, 600, 800, 1200, and 1600 µmol m−2 s−1) with 30 s duration; a 0.8 s saturating pulse was applied between each interval. The maximum quantum yield of PSII was given by the ratio Fv/Fm, where Fv = Fm – F0 (Schreiber et al., 1995). Effective quantum yield of PSII during actinic illumination (ΦPSII) was determined at 200 µmol m−2 s−1 following Genty et al. (1989). Maximal quenching owing to nonphotochemical dissipation of absorbed light energy (NPQmax) was determined at 1600 µmol m−2 s−1 (for further details see Arróniz-Crespo et al., 2006). A constant temperature of 10°C was maintained by taking all measurements inside a controlled environment chamber.

Three subreplicates per plot were measured for each physiological variable and the mean value of these was used for each plot.

Statistical analyses

Overall effects of long-term N and P treatments (i.e. excluding the recovery plots) were determined with two-way anova. Data were square-root or log10 + 1 transformed where necessary. Differences between the control plots and each treatment plot were determined by Dunnett's multiple comparison tests.

In cases where significant treatment effects were found, we then tested for recovery. Recovery is defined as a significant change in recovery plots compared with continually treated plots with the direction of change being towards control (0N plot) values. For this analysis, paired t-tests (pairing the ‘continually-treated’ and ‘recovery’ side of each plot) were run for each individual treatment to determine under which treatment recovery occurred. Paired t-tests on control (0N) plots revealed no significant differences between the continually treated control plots (distilled water only) and recovery control plots (no distilled water) for any of the variables measured for either species. We therefore make no further reference to 0N plots when discussing ‘recovery’. All statistical procedures were performed with SPSS 14.0.1 for Windows (SPSS Inc., Chicago, IL).


Percentage cover and sclerophylly

Nitrogen (N) treatments reduced the percentage cover of both bryophytes with up to 91% loss of R. squarrosus under the +14N treatment (P < 0.05). Phosphorus additions also reduced percentage cover of P. purum overall (Fig. 1a) (anova P < 0.01) but not R. squarrosus (Fig. 1b). No significant recovery of % cover was observed in either species.

Figure 1.

Percentage cover (a,b) and sclerophylly index (c,d) of Pseudoscleropodium purum and Rhytidiadelphus squarrosus from six different treatments: ambient nitrogen (N) (0N), enhanced N deposition at rates of 3.5 (3.5N) and 14 (14N) g N m−2 yr−1 alone and with phosphorus supplementation (0N + P, 3.5N + P and 14N + P) after either 11 yr 4 months of treatment (black bars for +N and grey bars for +P) or 9 yr 6 months of treatment plus 22 months recovery (open bars). Mean values (± SE) of three replicate plots are represented. Significant differences between the controls and the treated plots (*, P < 0.05; **, P < 0.01; ***, P < 0.001) were evaluated with Dunnett's tests. Black arrows indicate significant differences between the treated and recovery side of each plot (paired t-tests) (i.e. significant recovery).

Nitrogen additions reduced sclerophylly of P. purum (anova P < 0.001) but not R. squarrosus (Fig. 1c,d) while P additions did not effect sclerophylly in either species. A significant N × P interaction in R. squarrosus, however, suggested P may increase the sclerophylly of the plant when also combined with +N (anova N × P, P < 0.05).

Sclerophylly of P. purum showed significant recovery in formerly +14N and +14NP plots.

Tissue chemistry

Simulated N deposition significantly increased shoot N and N : P ratios in P. purum overall (anova P < 0.01 and P < 0.001, respectively) by up to 1.9- and 2.4-fold under +14N treatments (Fig. 2a,e) while P additions significantly increased shoot P and reduced N : P ratios (anova P < 0.001 for both) (Fig. 2c,e). Significant recovery was seen in P. purum shoot N in formerly +14N plots (P < 0.001) (Fig. 2a); this also occurred in shoot P in 0N + P and +14N + P plots (P < 0.05 for both) (Fig. 2c) and in shoot N : P in +3.5N plots (P < 0.05) (Fig. 2e).

Figure 2.

Nitrogen (N), phosphorus (P) and potassium (K) concentrations and N : P ratios of Pseudoscleropodium purum (a,c,e,f) and Rhytidiadelphus squarrosus (b,d,f,h). Treatments: ambient N (0N), enhanced N deposition at rates of 3.5 (3.5N) and 14 (14N) g N m−2 yr−1 alone and with P supplementation (0N + P, 3.5N + P and 14N + P) after either 11 yr 4 months of treatment (black bars for +N and grey bars for +P) or 9 yr 6 months of treatment plus 22 months recovery (open bars). Mean values (± SE) of three replicate plots are represented. Significant differences between the controls and the treated plots (*, P < 0.05; **, P < 0.01; ***, P < 0.001) were evaluated with Dunnett's tests. Black arrows indicate significant differences between the treated and recovery side of each plot (paired t-tests) (i.e. significant recovery).

By contrast, R. squarrosus shoot N, P and N : P ratios did not respond to +N treatments, but P additions did increase tissue P and reduce N : P (anova P < 0.001 for both) (Fig. 2b,d,f). Shoot P showed significant recovery in formerly +3.5N + P and +14N + P plots (P < 0.05 for both) (Fig. 2d).

Shoot K concentrations in R. squarrosus were reduced by +N treatments (anova P < 0.05) (Fig. 2h) with K concentrations in +14N plots being 68% less than in 0N controls (P < 0.001). Changes in shoot K concentrations were not detected in P. purum (Fig. 2g). In neither species was a recovery of shoot K detected (Fig. 2g,h).

Enzyme activities

Enhanced N deposition increased surface phosphomonoesterase activity (PME) in both species (anova P < 0.001 for P. purum and P < 0.01 for R. squarrosus) by up to two- and three-fold under +3.5N and +14N treatments (Fig. 3a,b) (P < 0.05 for both). In contrast, +P treatments significantly reduced PME activity in both species by c. 73% overall (anova P < 0.001 for both species) and a significant N × P interaction in P. purum (P < 0.001) suggested that +P treatments not only reduced PME activity but also dampened the magnitude of the +N effect.

Figure 3.

Surface phosphomonoesterase activity (p-NP release) (a,b) and induced nitrate reductase activity (NR) (c,d) of Pseudoscleropodium purum and Rhytidiadelphus squarrosus. Treatments: ambient nitrogen (N) (0N), enhanced N deposition at rates of 3.5 (3.5N) and 14 (14N) g N m−2 yr−1 alone and with phosphorus (P) supplementation (0N + P, 3.5N + P and 14N + P) after either 11 yr 4 months of treatment (black bars for +N and grey bars for +P) or 9 yr 6 months treatment plus 22 months recovery (open bars). Mean values (± SE) of three replicate plots are represented. Significant differences between the controls and the treated plots (*, P < 0.05; **, P < 0.01; ***, P < 0.001) were evaluated with Dunnett's tests. Black arrows indicate significant differences between the treated and recovery side of each plot (paired t-tests) (i.e. significant recovery).

Pseudoscleropodium purum showed significant recovery of PME (Fig. 3a) with a 21% and 36% decline towards control values in formerly +3.5N and +14N plots (P < 0.05 for both), respectively. No recovery from +P treatments was detected in P. purum and R. squarrosus shown no significant recovery from either +N or +P treatments (Fig. 3b).

Inducible nitrate reductase (NR) of R. squarrosus was 42% lower in both +3.5N and +14N plots compared with controls (Fig. 3d) (P < 0.05 for both). A significant N × P interaction in both species resulted from +N reducing NR more when combined with the +P treatment. The +P treatments caused a significant increase from 0N control values in NR, by 222% in 0N + P for P. purum and 214% and 182% (0N + P and +3.5NP respectively) in R. squarrosus.

The NR of R. squarrosus showed significant recovery in former +3.5N and +14N plots (Fig. 3d) with an increase to around control values, but no recovery of NR was observed in P. purum (Fig. 3c).

Pigment composition

As a result of the considerable reduction in cover, there was too little biomass of R. squarrosus to harvest for pigment analysis in the +14N plots.

Simulated N deposition significantly increased total chlorophyll (both chl a and b) in P. purum (anova P < 0.05) (Fig. 4a) and reduced the Chl a/b ratios in P. purum and R. squarrosus (anova P < 0.01 and P < 0.05, respectively) (Fig. 4c,d). The N deposition also increased the (neoxanthin + lutein)/β-carotene ratio in both species (anova P < 0.05 both species) (Fig. 4e,f) and reduced the DES in R. squarrosus (anova P < 0.05) (Fig. 4). In contrast to N, the +P treatment significantly reduced the DES in P. purum (anova P < 0.05).

Figure 4.

Total chlorophyll concentration, chlorophyll a/b ratio, (neoxanthin + lutein)/β-carotene ratio and de-epoxidation of the xanthophyll cycle pool (DES) of Pseudoscleropodium purum (a,c,e,g) and Rhytidiadelphus squarrosus (b,d,f,h). Treatments: ambient nitrogen (N) (0N), enhanced N deposition at rates of 3.5 (3.5N) and 14 (14N) g N m−2 yr−1 alone and with phosphorus (P) supplementation (0N + P, 3.5N + P and 14N + P) after either 11 yr 4 months of treatment (black bars for +N and grey bars for +P) or 9 yr 6 months of treatment plus 22 months recovery (open bars). Mean values (±SE) of three replicate plots are represented. Significant differences between the controls and the treated plots (*, P < 0.05; **, P < 0.01; ***, P < 0.001) were evaluated with Dunnett's tests. Black arrows indicate significant differences between the treated and recovery side of each plot (paired t-tests) (i.e. significant recovery).

Pseudoscleropodium purum showed significant recovery of chlorophyll concentration in formerly +14N plots (P < 0.05), Chl a/b ratio (P < 0.01) and (neoxanthin + lutein)/β-carotene ratio (in +14N plots, P < 0.05) (Fig. 4a,c,e). Significant recovery of R. squarrosus Chl a/b ratio and (neoxanthin + lutein)/β-carotene ratio was seen in formerly +3.5N plots (P < 0.05 for both) and of DES in formerly +3.5N and +14N + P plots (P < 0.05 and P < 0.01, respectively) (Fig. 4d,f,h).

Chlorophyll fluorescence

Simulated N deposition significantly reduced Fv/Fm in both species (anova P < 0.01 for P. purum and P < 0.05 for R. squarrosus) (Fig. 5a,b) and ΦPSII (P < 0.001 for P. purum and P < 0.05 for R. squarrosus) (Fig. 5c,d) and NPQmax (P < 0.01 for both species) (Fig. 5e,f). Fv/Fm was reduced most in P. purum, from 0.73 under 0N to 0.69 and 0.68 under +3.5N and +14N, respectively (P < 0.01, P < 0.001), while in R. squarrosus this reduction was from 0.73 to 0.70 (P < 0.05) under +14N. The ΦPSII was also seen to be more sensitive in P. purum, falling by 23% and 55% under the +3.5N and +14N treatments (P < 0.01 and P < 0.001, respectively), but by only 27% in R. squarrosus under +14N (P < 0.05). In contrast, NPQmax showed the greatest decline in R. squarrosus compared with P. purum being reduced in all the treatments in the former species and in +14N (with and without P) in the latter species.

Figure 5.

Maximum (Fv/Fm) and effective (ΦPSII) quantum yield of PSII and maximum nonphotochemical quenching (NPQmax) of Pseudoscleropodium purum and Rhytidiadelphus squarrosus. Treatments: ambient nitrogen (N) (0N), enhanced N deposition at rates of 3.5 (3.5N) and 14 (14N) g N m−2 yr−1 alone and with phosphorus (P) supplementation (0N + P, 3.5N + P and 14N + P) after either 11 yr 4 months treatment (black bars for +N and grey bars for +P) or 9 yr 6 months treatment plus 22 months recovery (open bars). Mean values (± SE) of three replicate plots are represented. Significant differences between the controls and the treated plots (*, P < 0.05; **, P < 0.01; ***, P < 0.001) were evaluated with Dunnett's tests. Black arrows indicate significant differences between the treated and recovery side of each plot (paired t-tests) (i.e. significant recovery).

The +P treatment increased Fv/Fm and ΦPSII in P. purum (P < 0.01 and P < 0.001 respectively) and decreased NPQmax in R. squarrosus (P < 0.05).

Most fluorescence parameters tended to show a return towards control values in formerly N-treated plots but statistically significant recoveries were limited to Fv/Fm and ΦPSII of P. purum (in +3.5N plots +14N plots respectively) and to NPQmax of R. squarrosus in +14N + P plots.


This study has undertaken a wide-ranging analysis of multiple physiological variables in two widely distributed bryophytes to N deposition impacts, P additions and recovery.

We found that both species were highly sensitive to simulated enhanced N deposition, showing major and significant reductions in % cover at all levels of N addition. A number of key physiological variables were also shown to be highly sensitive to enhanced N deposition, providing insight into possible mechanisms causing the decline in abundance and which may have value as sensitive indicators of bryophyte responses to N loading. Importantly, while 22 months cessation of N treatments resulted in no clear recovery of cover, some physiological variables – particularly PME, shoot N, N : P and sclerophylly index (SI) – show clear recovery in P. purum, revealing their potential to provide sensitive early indicators of recovery from historic N loading.

Percentage cover and sclerophylly

The decline in abundance of the two dominant bryophytes within this acidic grassland supports earlier work undertaken on a neighbouring experiment in the same ecosystem (Lee & Caporn, 1998; Carroll et al., 2003). Our study further found important interspecific differences in sensitivity since R. squarrosus lost c. twofold more cover than P. purum. As discussed later, our physiological measurements suggest the declines in cover may be partially driven by increased P stress, increased fragility (reduced SI) and a decline in maximum potential photosynthetic rates despite investment of N in tissue leading to increased chlorophyll concentrations.

Interspecific differences in response to P addition (i.e. an alleviation of P limitation) were also seen with +P treatments, further reducing P. purum cover while having little impact on R. squarrosus. While P amendment has been shown to benefit bryophytes in some N addition studies (Limpens et al., 2004; Pilkington et al., 2007), our study suggests that such benefits may not occur in slow-growing species such as P. purum, possibly because they may become outcompeted by species that are not adversely affected by, or perhaps benefit from, P amendment.

The SI in P. purum showed a similar pattern of response to N deposition as percentage cover but also showed recovery that could not be detected in percentage cover. Reduction of cell wall thickness has been observed in some bryophytes when grown in a nutrient-rich medium (Voth, 1943) and may partly explain the reduction of SI in response to N deposition seen in our study. Alternatively, because surface area is measured on prostrate shoots and leaves overlap, the decreased SI may reflect a lowering of leaf density along the shoot (pers. obs.) resulting in less leaf overlap and a lower density of tissue per unit surface area. Both mechanisms of reduced SI, however, would suggest an increase in fragility (i.e. less sclerophylly), which could contribute to the decline in cover of P. purum.

Overall though, since SI of R. squarrosus was, in contrast, found to be unresponsive to N deposition, we cannot suggest SI as a good indicator of N deposition impacts in bryophytes in general.

Tissue N, K and P

In P. purum, increased N deposition resulted in increased tissue N and tissue N : P ratios whereas R. squarrosus showed no such changes. This latter absence of response has also been observed in R. squarrosus along an N deposition gradient across the UK (Stevens et al., 2006). These findings suggest tissue N may not be a robust indicator of N deposition impacts in bryophytes generally. Higher accumulation of N in P. purum compared with R. squarrosus was also observed by Solga & Frahm (2006) in a short-term fertilization experiment. Possibly a higher desiccation tolerance and opportunistic strategy for nutrient uptake in P. purum might confer a more direct relationship between N supply and N uptake (Bates, 1997). Furthermore, since R. squarrosus was already N saturated under +3.5N (as suggested by the significant decrease in NR), this would limit its N assimilation capacity and increase in tissue N. In this case, while the presence of rhizoids in R. squarrosus does not seem to be exacerbating increases in tissue N (by increasing uptake from the soil) they may be contributing to this moss becoming readily N saturated even under the lower rates of N deposition.

The reduction in shoot K concentration in R. squarrosus resulting from increased N deposition may result from exchange competition between inline image and K+ ions or, more probably, by a direct toxic effect of N that leads to leakage of base cations through damaged cell membranes (Pearce et al., 2003). Direct toxic effects of N could also be exacerbated where decreased NR activity limits the bryophyte's capacity to assimilate N (Pearce et al., 2003). Since, in contrast, significant reduction in K was not observed in P. purum, this might indicate higher tolerance of this species to N deposition, which is consistent with the lesser reduction in % cover of this species compared with R. squarrosus.

Phosphomonoesterase and nitrate reductase

Our work reports the first evidence that PME activity is significantly stimulated in bryophytes under increased N loads – a finding consistent with similar responses seen in higher plant roots (Phoenix et al., 2004). Since this enzyme may increase the availability of phosphate (Turner et al., 2001), greater stimulation of PME by N loads in P. purum compared with R. squarrosus may also confer greater ability of the former species to enhance P uptake (and so reduce P limitation) under N enrichment. This response might also facilitate the higher assimilation of N in P. purum compared with R. squarrosus since the potential for bryophytes to use N can depend on the availability of other nutrients, especially P and K (Carfrae et al., 2007).

Reductions in induced NR under increased N loads have been well documented and are known to indicate reduced demand for N and the onset of N saturation since its activity is end-product inhibited (Glime, 2007). Interestingly, the magnitude of change in NR was much smaller than for PME. This difference may result if the assay for PME is more sensitive than that for NR, but more likely it suggests that exacerbation of P limitation is a more important driver of bryophyte responses to N deposition than N saturation per se.

In contrast to N loads, additions of P considerably reduced PME and increased NR activity in both species (changes that are indicative of a shift from P to N limitation under P amendment). Since decreases in shoot K caused by N loading in R. squarrosus disappeared with the addition of P, this supports the beneficial effect of P in this species: a possible explanation being the reduction of N toxicity through P amendments facilitating N assimilation.

The sensitivity and magnitude of response of PME and NR suggest that these have value as indicators of N deposition impacts. They may also provide good indicators of recovery since PME activity showed a marked return towards control levels in P. purum in N recovery plots (and some return was also seen in R. squarrosus– though not significantly so) and NR showed apparent recovery in R. squarrosus (although this should be viewed with caution since ‘recovery’ of NR was also seen in control plots suggesting the shift may be an artefact of spatial or temporal variation). Overall, while NR activity has previously been used as good indicator of N deposition impacts (Woodin & Lee, 1987; Pearce et al., 2003; Pitcairn et al., 2006), our work suggests that PME may actually provide a clearer (and possibly more sensitive) indicator of both N deposition impacts and recovery than NR.

Pigments and chlorophyll fluorescence

Increases in total chlorophyll (both chl a and b) in P. purum under enhanced N deposition suggests an investment of accumulated N in these pigments. The accompanying reductions in the Chl a/b ratio and increases in the (neoxanthin + lutein)/β-carotene ratio show that increased N loads alter the composition of pigment protein complexes in the thylakoid membranes and increase the relative proportion of light harvesting complexes compared with reaction centres. Interestingly, the increased pigment concentrations were associated with a significant decrease in the efficiency of PSII as both Fv/Fm and ΦPSII decreased under increased N deposition. An indirect shade effect of higher plants could explain the reduction of ΦPSII and the observed changes in pigment composition (Lambers et al., 1998). However, the N dose-dependent response and fast recovery detected in these variables (Fv/Fm, ΦPSII, total Chl, and Chl a/b and (neoxanthin + lutein)/β-carotene ratios) suggest that direct N loading is the main driver and not higher plant shading: in these P-limited grasslands, increased productivity of higher plants either does not occur or occurs little under increased N loads (and therefore also cannot show rapid recovery) (Carroll et al., 2003; Phoenix et al., 2003).

Overall, these changes suggest impairment of maximum potential photosynthetic rates, particularly as indicated by evidence of photo-inhibition (decline in Fv/Fm), and decline in energy utilization in photosynthesis (declines in ΦPSII). Since these responses are seen in both species, impairment of photosynthesis and hence growth, may be an important driver of decline in bryophyte cover.

In contrast to that direct effect, increased higher plant productivity that occurs with +P amendment in this grassland (Phoenix et al., 2003) would cause a shading effect of higher plants explaining the lower DES and NPQmax of P. purum in +P amended plots and possibly contributing to the reduction of cover of this moss under +P treatments.

In the case of R. squarrosus it was more difficult to assess the impacts of N loading on pigment composition because of limited plant material in the +14N plots. However, significant +N effects were still detected in the Chl a/b and (neoxanthin + lutein)/β-carotene ratios, following a similar pattern to that seen in P. purum. In R. squarrosus it was interesting to observe that total chlorophyll concentration seems to increase under combined +N +P treatments suggesting that +P addition might facilitate N assimilation as mentioned previously for this species: this is also consistent with increases in NR under +N +P treatments compared with +N only treatment.

Again, in contrast to the lack of recovery from N loading seen with percentage cover, a rapid recovery was observed in Fv/Fm, ΦPSII and total chlorophyll concentration in P. purum, NPQmax in R. squarrosus and Chl a/b and (neoxanthin + lutein)/β-carotene ratios in both species. Photosynthetic pigments and nondisruptive techniques such as chlorophyll fluorescence could be good complementary methods for early and sensitive indication of recovery from N pollution impacts and could be used to support other approaches such as PME measurements.


This study has undertaken an extensive assessment of bryophyte physiological responses to simulated enhanced N deposition, P amendment and recovery. In doing so, we have elucidated some of the drivers of sensitivity and changes in abundance of two common grassland bryophytes, as well as suggesting physiological parameters that may prove useful as sensitive indicators of N deposition impacts and recovery in national air-pollution monitoring schemes.

This work found P. purum to be more tolerant of simulated enhanced N deposition than R. squarrosus, a difference that may be attributable to a higher tolerance to P stress in P. purum, enabling this bryophyte to accumulate and invest the extra N supply in its photosynthetic apparatus (see the Supplementary Material, Table S1 for summary of interspecific differences in sensitivity). Intriguingly, since this greater tolerance of P. purum is conferred by a greater capacity to respond to N loading than R. squarrosus (in showing the greater physiological responses), it means P. purum is the better bioindicator species of N pollution impacts and recovery (and certainly over the range of N deposition rates we have tested). We also highlight that while 22 months cessation of N treatments resulted in no clear recovery of cover, many physiological variables did show recovery, in particular, changes in PME, N : P ratio and SI in P. purum and NR in R. squarrosus. Overall, the physiological variable of greatest value as a sensitive indicator of N deposition impacts and recovery in P. purum is PME, with tissue N : P ratios, Fv/Fm, and pigment analysis being useful as additional corroborating pollution indicators.

Validation of these variables as indicators of N deposition impacts and recovery in landscape and national scale surveys warrants further attention.


We are grateful to the Royal Society (Research Grant 2006/R1), NERC (grant NE/D00036X/1) and DEFRA (Terrestrial Umbrella Project) for their financial support. M.A-C. benefited from a postdoctoral grant of the Ministerio de Educación y Ciencia of Spain and the Fondos FEDER (Project CGL2005-02663). G.K.P. is funded by an RCUK academic fellowship. We are grateful to Ben le Bas and Natural England for allowing access to the field site and Professor Paul Quick for allowing the use of his spectrophotometer. We thank Dr Mike Proctor for his useful comments during the development of the project and Dr Duncan Cameron for helpful comments on the manuscript.