Advertisement

Hydrobiologia

, Volume 674, Issue 1, pp 41–50 | Cite as

Tracing Si–N–P ecosystem-pathways: is relative uptake in riparian vegetation influenced by soil waterlogging, mowing management and species diversity?

  • Eric Struyf
  • Wiktor Kotowski
  • Sander Jacobs
  • Stefan Van Damme
  • Kris Bal
  • Wout Opdekamp
  • Hans Backx
  • Dimitri Van Pelt
  • Patrick Meire
Open Access
Wetland Restoration

Abstract

Despite the growing concern about the importance of silicon (Si) in controlling ecological processes in aquatic ecosystems, little is known about its processing in riparian vegetation, especially compared to nitrogen (N) and phosphorus (P). We present experimental evidence that relative plant uptake of N and P compared to Si in riparian vegetation is dependent on mowing practices, water-logging and species composition. Results are obtained from a controlled and replicated mesocosm experiment, with a full-factorial design of soil water logging and mowing management. In our experiments, the Si excluding species Plantago lanceolata was dominant in the mown and non-waterlogged treatments, while Si accumulating meadow grasses and Phalaris arundinacea dominated the waterlogged treatments. Although species composition, management and soil moisture interacted strongly in their effect on relative Si:N and Si:P uptake ratios, the uptake of N to P remained virtually unchanged over the different treatments. Our study sheds new light on the impact of riparian wetland ecosystems on nutrient transport to rivers. It indicates that it is essential to include Si in future studies of the impact of riparian vegetation on nutrient transport, as these are often implemented as a measure to moderate excessive N and P inputs.

Keywords

Riparian meadows Riparian wetlands Silicon Nitrogen Phosphorus Vegetation Eutrophication 

Introduction

The uptake, storage and recycling of silicon (Si) in ecosystem biomass and soils are now recognized as important components of silicon transport through the terrestrial environment (Conley, 2002; Derry et al., 2005; Street-Perrott & Barker, 2008; Struyf & Conley, 2009), but the relative importance of the biological buffer in different ecosystems, and the impact of environmental forcing factors on the efficiency of storage, uptake and recycling still remains poorly quantified. Si is a key nutrient in determining the species composition of aquatic and coastal phytoplankton communities (Cloern, 2001). The commonly observed effect of increased anthropogenic inputs of nitrogen (N) and phosphorus (P) to aquatic ecosystems is a shift in the coastal phytoplankton community to one dominated by dinoflagellates (Lotze et al., 2006), caused by dissolved Si (DSi) limitation of diatoms (Conley et al., 1993). A lack of DSi can lead to a less efficient food web structure as well as to hypoxia and harmful blooms of toxic algae (e.g. Anderson et al., 2002). Silicon also plays an essential role in the global carbon (C) cycle. A continuous import of Si from the terrestrial environment is necessary to maintain the biological C pump in the ocean (Tréguer & Pondaven, 2000), and mineral weathering of silicate minerals is one of the major terrestrial sinks for carbon dioxide (CO2) (Berner, 1992). The poor quantification of the biological Si buffer in ecosystems forms a prime challenge in constraining the silicon cycle related C sinks.

Wetlands and meadows are prime candidates to form a hot-spot in the biological Si buffer, as they can accumulate large amounts of biogenic amorphous Si (ASi) (Blecker et al., 2006; Struyf & Conley, 2009) in the form of plant phytoliths, diatoms and sponge spicules. Riparian meadows are well-established buffers in the N and P cycle. They affect the output of these nutrients to rivers through permanent or temporary uptake in vegetation, detritus and soils and through transformation processes, e.g. removal by denitrification (e.g. Güsewell, 2005; Olde Venterink et al., 2006). Riparian wetlands are often (re)established as a countermeasure against excessive N and P export from agricultural catchments. Up to now however, little research has focused on the influence of ecosystem characteristics on Si processing in riparian meadows or wetlands. Management has been observed to potentially interfere with the storage of ASi near the surface of riparian wetlands (Struyf et al., 2009), but experimental evidence is currently lacking (Struyf & Conley, 2009).

We analyzed uptake of Si by riparian vegetation in reference to N and P in mesocosm communities grown for one season under contrasting moisture and management conditions. Earlier experiments have focused on N, P and often K and C, while failing to include Si. Our study sheds new light on the impact of riparian wetland ecosystems on nutrient transport to rivers, and is new evidence of the importance of the role of vegetation in terrestrial Si fluxes.

Materials and methods

Mesocosms

We established a full-factorial mesocosm experiment with soil water-logging and summer mowing as factors, each of them in two levels (waterlogged, anoxic vs. non-waterlogged, oxic and mown vs. unmown). The experimental set-up is described in Kotowski et al. (2010). Briefly, mesocosms were composed of PVC containers of 91 × 111 cm surface and 61 cm height. Outflow taps located at the bottom of side walls connected to PVC pipes allowed for the control of the water level. Containers were filled with 20 cm of drainage substrate (Argex clay aggregate, Argex NV, Zwijndrecht, Belgium) separated by geo-textile from a 30 cm cover of a soil mixture containing 3:1 volumetric parts of fen peat and alluvial clay. Mixing fen peat and clay ensured that the soil would have high water potential and mineral content (clay) and at the same time, high organic matter content (peat) necessary to produce reduced conditions after water-logging. Presence of peat in the soil also allowed for easier percolation of water under non-water logged conditions. To maintain eutrophic conditions throughout the experiment, an initial amount of 168 g N, 192 g P, 216 g potassium (K) and 24 g magnesium (Mg) per m3 soil in an easily soluble form (PG-mix Floranid, COMPO Benelux nv, Deinze, Belgium), and a slow release fertilizer (6 months release time; Osmocote, SCOTTS Professional, Sint-Niklaas, Belgium) containing 300 g N, 180 g P and 180 g K per m3 soil were added. These values are at the upper limit of annual nutrient input in European floodplains (Olde Venterink et al., 2006).

Forty mesocosms were placed in a block of 4 × 10. Water was supplied by sprinklers installed 30 cm above each container once (spring, autumn) or twice (summer) a day. Watering time depended on the time of the year (between 5 and 25 min per day) and specifically ensured that plants were not limited by water shortage. Excessive water could flow out of the system through the outflow pipe. In anoxic soils, the water level was held constant at 2–5 cm below soil surface, while the outflow was levelled at the bottom of the containers in the oxic soils.

The species used in the experiment were typical of Central European floodplains. We carefully selected species as to have a similar representation of typical fen, wet meadow and mesic meadow species and to assure that each of these groups contained species from various functional types and growth forms (Supplementary material). A seed mixture of 37 species was introduced on the 15th of May 2006 in a density of approximately 1,000 seeds m−2 species−1. Soil moisture measurements at 5 cm depth carried out in mid June between two watering treatments confirmed differences between the waterlogged and the non-waterlogged treatments (soil moisture: 54.2% ± 12.1 and 32.4% ± 3.5, respectively, by weight, t test P < 0.001).

Sampling

The above-ground biomass of the 20 mesocosms subjected to the summer mowing treatment was harvested at the end of June 2006. Biomass was sorted to species level, except for the meadow grasses which were grouped. Dry biomass of all species was determined separately (except for the meadow grasses) in each mesocosm after 72 h of drying at 70°C. At the end of the experiment (September 2006), the above-ground biomass of all 40 mesocosms was harvested and the sorting and drying was carried out as described above.

Nutrient analysis

N, P and Si concentrations in plant biomass for all established species were determined. If a species established in more than seven mesocosms with the same combination of mowing and aeration treatment, seven randomly chosen replicates were analyzed for their N, P and Si content.

After drying, the plant material was grounded with a motor mill and analyzed for N, P and Si content. ASi content was analyzed through incubation of 25 mg of ground, homogenized plant material for 4 h in 0.1 M Na2CO3 at 80°C (DeMaster, 1981). Extracted dissolved silicon was subsequently analyzed spectrophotometrically on a Thermo IRIS ICP (Inductively Coupled Plasmaspectrophotometer). N and P content were analyzed according to Walinga et al. (1989) on a colorimetric segmented flow analyzer.

Nutrient uptake at the vegetation level

To allow the analysis of nutrient uptake at community level during the experiment (May–September 2006), the dry weight of every species was multiplied with the average N, P and Si content observed for the species in a certain treatment. The total N, P and Si uptake in biomass for every mesocosm was then calculated by summation of the calculated total nutrient uptake in all species. For the mown treatments, this is the summation of the May–June and July–September uptake.

Statistical analyses

A one-way ANOVA analysis was used to comparing the treatments for different variables (Biomass, N, P and Si uptake, N:P ratio, N:Si and P:Si). Classification of the treatments was performed with Post-hoc Scheffe tests. All analyses were performed in SPSS (Statistical Package for the Social Sciences) 14.0 (SPSS, 2005).

Results

Vegetation patterns and biomass

In total, 26 species established in the mesocosms but only 14 were present in sufficient amounts to allow for chemical analyses (see Kotowski et al., 2010 for detailed information on species recruitment in each treatment). Three meadow grasses established (i.e. Lolium perenne L., Poa pratensis L., and Alopecurus pratensis L.): these were grouped as one taxon within the analysis. Another grass species, Phalaris arundinacea L. was considered separately, as it is a typical wetland grass, while the other grouped species are typical meadow grasses. As such, 11 species and one group-taxon were included in the analysis: “meadow grasses (L. perenne, P. pratensis and A. pratensis)”, Achillea millefolium L., Centaurea jacea L., Cirsium oleraceum L. Scop., Daucus carota L., Epilobium hirsutum L., Leontodon autumnalis L., Lycopus europaeus L., Lythrum salicaria L., P. arundinacea, Plantago lanceolata L. and Valeriana officinalis L..

The total above-ground biomass was higher in the non-waterlogged, oxic treatments (Table 1). Biomass in oxic treatments was dominated by P. lanceolata, and to a lesser extent, by P. arundinacea, D. carota and A. millefolium (Fig. 1). Biomass in the waterlogged, anoxic treatments was dominated by meadow grasses and P. lanceolata, and to a lesser extent by P. arundinacea (Fig. 1). A clear shift of the dominant species was apparent between anoxic and oxic treatments, with meadow grasses dominating the anoxic treatments and P. lanceolata the oxic treatments (Fig. 1). More than 80% of the decrease in total biomass under anoxic conditions could be attributed to P. lanceolata, with a strong decline in biomass, both in mown (−1,000 g dry weight m−2 of P. lanceolata biomass on average) and unmown mesocosms (−1,360 g dry weight m−2 on average). Given the importance of P. lanceolata for total biomass production, we also tested for differences in biomass excluding P. lanceolata. Although a marginally significant effect of treatment on biomass without P. lanceolata biomass existed (P < 0.04), the post-hoc Scheffe tests did not separate any of the treatments individually (Table 1).
Table 1

Results for a one-way Anova analysis comparing the four treatments for different variables: biomass, nitrogen (N), phosphorus (P), silicon (Si) uptake, and weight ratios of N:P, N:Si and P:Si

Variable

F value and P value

Effect

Total biomass

 Biomass

F 3.39 = 26.46; P < 0.001

ONM = OM > AM = ANM

 N uptake

F 3.39 = 60.504; P < 0.001

OM > ONM > AM = ANM

 P uptake

F 3.39 = 45.90; P < 0.001

OM > ONM > AM = ANM

 Si uptake

F 3.39 = 18.77; P < 0.001

ONM = OM > AM = ANM

 N:P ratio

F 3.39 = 140.963; P < 0.001

OM > ONM > AM > ANM

 N:Si ratio

F 3.39 = 15.826; P < 0.001

AM = OM > ONM = ANM

 P:Si ratio

F 3.39 = 13.674; P < 0.001

AM = OM > ONM = ANM

Without Plantago

 Biomass

F 3.39 = 3.107; P < 0.04

ONM = OM = AM = ANM

 N uptake

F 3.39 = 1.45; P > 0.2

ONM = OM = AM = ANM

 P uptake

F 3.39 = 1.30; P > 0.2

ONM = OM = AM = ANM

 Si uptake

F 3.39 = 12.84; P < 0.001

ONM = OM >AM = ANM

 N:P ratio

F 3.39 = 30.076; P < 0.001

OM > ONM = AM = ANM

 N:Si ratio

F 3.39 = 75.126; P < 0.001

AM > ANM > OM = ONM

 P:Si ratio

F 3.39 = 78.514; P < 0.001

AM > ANM > OM = ONM

The result for the full Anova test is shown (P value for significance, F value as the critical value extracted from the f-distribution with 3 and 39 degrees of freedom), as well as the classification of the treatments according to Post-hoc Scheffe tests (OM oxic, mown; ONM oxic, unmown; AM anoxic, mown; ANM anoxic, unmown). The analysis was repeated for the full vegetation, and for vegetation excluding Plantago lanceolata, the main silicon excluder in the analysis

Fig. 1

Contribution of different species to total biomass in the different treatments (averaged, n = 10 for every treatment)

Nutrient concentrations at the species level

The plant nutrient concentration among the different replicates only showed small variability (Table 2). Results for nutrient contents will focus on the main dominant species D. carota, A. millefolium, “meadow grasses”, P. arundinacea and P. lanceolata, which represented over 90% of nutrient uptake. Average biomass for the dominant species, and average N, P and Si concentrations in the plant tissue is summarized in Table 2. P. lanceolata and D. carota had low N, P and Si contents compared to the other dominant species. The species in the anoxic treatment generally contained less N, P and Si compared to the oxic treatments. Highest relative differences were observed for Si content in P. arundinacea, meadow grasses and A. millefolium, which had more than twice the relative Si content in oxic compared to anoxic treatments.
Table 2

Biomass (BM), silicon (Si), nitrogen (N) and phosphorus (P) contents for all dominant species, sampled in June (first mowing) and September (definitive sampling)

 

Treatment

Oxic, unmown

Oxic, mown

Anoxic, unmown

Anoxic, mown

September

BM

SE

Si

SE

n

BM

SE

Si

SE

n

BM

SE

Si

SE

n

BM

SE

Si

SE

n

Meadow grasses

97

23

8.6

1.3

7

37

22

6.6

0.7

7

575

20

3.3

0.2

7

160

28

3.0

0.1

7

Achillea millefolium

267

40

7.9

0.9

7

18

4

10.2

1.7

6

7

1

4.0

0.5

6

3

1

3.2

0.4

3

Daucus carota

227

48

1.9

0.9

7

81

21

0.9

0.4

7

16

1

0.5

0.1

7

18

9

0.7

0.4

7

Phalrais arundinacea

378

72

10.7

1.5

7

195

41

9.8

2.3

7

245

43

4.3

0.6

7

143

32

4.7

2.4

7

Plantago lanceolata

1090

48

0.9

0.2

6

760

88

0.5

0.1

7

157

31

0.5

0.1

6

229

80

0.5

0.1

7

 

N

SE

P

SE

n

N

SE

P

SE

n

N

SE

P

SE

n

N

SE

P

SE

n

Meadow grasses

19.0

0.8

3.7

0.1

7

23.3

2.8

5.1

0.2

7

14.7

0.6

3.2

0.2

7

18.5

2.0

3.9

0.4

7

Achillea millefolium

19.5

1.6

3.8

0.3

7

22.2

2.0

4.5

0.4

6

14.1

0.6

2.7

0.1

6

20.5

0.9

3.6

0.2

3

Daucus carota

14.3

1.6

3.4

0.4

7

19.3

1.5

4.1

0.3

7

12.9

0.5

3.1

0.2

7

15.0

2.1

3.2

0.2

7

Phalrais arundinacea

18.3

0.6

3.8

0.2

7

20.3

2.0

3.9

0.4

7

9.5

0.9

2.0

0.2

7

15.3

4.0

2.7

0.6

7

Plantago lanceolata

16,6

1.2

3.2

0.2

6

18.2

1.9

3.1

0.3

7

10.5

0.6

2.4

0.1

6

14.3

1.0

2.7

0.2

7

June

     

BM

SE

Si

SE

n

     

BM

SE

Si

SE

n

Meadow grasses

     

72

8

10.4

1.3

7

     

309

48

3.1

0.2

7

Achillea millefolium

     

139

21

13.0

1.9

7

     

7

2

4.0

0.5

6

Daucus carota

     

70

11

1.8

0.1

7

     

17

2

1.1

0.1

7

Phalrais arundinacea

     

23

5

12.9

2.0

7

     

45

7

3.8

0.3

7

Plantago lanceolata

     

888

70

0.6

0.1

7

     

150

29

0.4

0.1

7

      

N

SE

P

SE

n

     

N

SE

P

SE

n

Meadow grasses

     

31.0

0.8

5.7

0.2

7

     

19.7

0.7

4.5

0.2

7

Achillea millefolium

     

25.8

2.0

5.6

0.3

7

     

17.9

0.8

3.7

0.2

6

Daucus carota

     

23.3

1.8

4.3

0.2

7

     

13.7

0.7

2.9

0.4

7

Phalrais arundinacea

     

31.0

1.6

5.3

0.4

7

     

17.4

0.9

2.9

0.1

7

Plantago lanceolata

     

20.6

2.4

3.8

0.4

7

     

9.7

0.5

2.4

0.1

7

Biomass on dry weight basis (g m−2) is averaged over all 10 replicates for every treatment but for chemical nutrient analysis we only used maximally 7 (randomly chosen) replicates. Silicon (Si), N and P are indicated as relative contents in the plant tissue (mg g−1 dry weight). All indicated errors are standard errors (SE, n = number of replicates). Unmown treatments were not sampled in June

Nutrient uptake

Total uptake of nutrients (N, P and Si) in the oxic treatments was high compared to anoxic treatments (Fig. 2; Table 1). Under oxic circumstances, N and P uptake were higher in the mown treatments; this effect was not apparent for Si uptake (Table 1). The differences for N and P uptake were attributable to P. lanceolata biomass. Excluding this, no differences in total N and P uptake were apparent (Table 1, Fig. 2 lower panel). This was not the case for Si uptake; excluding P. lanceolata biomass had no effect on Si uptake (Table 1; Fig. 2) as there is almost no Si in P. lanceolata (Table 2).
Fig. 2

Total nitrogen (N), phosphorus (P) and silicon (Si) uptake in the vegetation of the different treatments. The top panels are for the total biomass, the lower panels for the biomass without P. lanceolata. The error bars indicate the standard deviation (OM oxic, mown; ONM oxic, unmown; ANM anoxic, mown; ANNM anoxic, unmown)

In mown treatments more N and P were taken up compared to Si than in unmown treatments, both under oxic and anoxic conditions (Fig. 3; Table 1). The relative uptake of Si dropped mainly as a result of increasing P. lanceolata abundance in the mown treatments (N:Si ratio in P. lanceolata ranged between 17 and 38, while, e.g. P. arundinacea N:Si ranged between 2 and 5). Although significant differences existed in N:P ratio across the treatments (Table 1), these differences were very small (Fig. 3).
Fig. 3

Nitrogen (N):phosphorus (P), P:silicon (Si), and N:Si weight ratios in the vegetation in the different treatments. The left panel is for the total biomass, the right panel for the biomass without P. lanceolata. The error bars indicate the standard deviation (OM oxic, mown; ONM oxic, unmown; ANM anoxic, mown; ANNM anoxic, unmown)

If we remove P. lanceolata from the analysis, all remaining dominant vegetation consists of Si accumulators. This had little effect on N:P ratio in the vegetation across the treatments (Fig. 3; Table 2), although small significant differences still existed (Table 2). Removing P. lanceolata from the analysis had almost no effect on the total Si content of plants, while the N and P taken up by the vegetation decreased drastically (Fig. 2). The N:Si and P:Si ratio in the Si accumulator (non-Plantago) biomass was mainly related to the water-logging in the soil (Table 1, Fig. 3), although mowing had a significant effect in the anoxic treatments (Table 1).

In summary, the N:P ratio only differed marginally over species composition, soil moisture and mowing conditions. In contrast, the ratio at which Si was taken up relative to the other nutrients was significantly different. Mowing enhanced occurrence of P. lanceolata, and as such decreased relative uptake of Si. In the accumulator biomass, relative uptake of Si to N and P was mainly related to water logging.

Discussion

Although abundant experimental studies have focused on N and P uptake in riparian vegetation, a controlled experiment which studies Si uptake (relative to N and P) in detail, under changing environmental conditions, is surprisingly lacking. Riparian meadows, of which our experimental units were small-scale versions, have a high potential to influence the Si input into rivers. Often dominated by Si-rich grasses or sedges, they are likely among the most active processors of Si of all terrestrial ecosystems (e.g. Blecker et al., 2006; Sommer et al., 2006; Struyf & Conley, 2009).

P concentrations in our experiment (the 4 dominant species contain on average 3.7 mg g−1 dry weight) were high compared to other studies, while N concentrations (18.7 mg g−1 dry weight) were high but not uncommon (Güsewell & Koerselman, 2002). Our experiment was, given the high supply of N and P through fertilisers and the relatively high concentrations of N and P in plant tissues, clearly representative for a set of riparian meadows in an eutrophied environment, with N less available compared to P (but neither should be considered limiting at measured concentrations). In such eutrophied environments, wetland restoration and protection is often implemented as a measure to reduce N and P export. It is a well-established concept that controlling N and P export upstream by implementing riparian vegetations is one of the most effective ways to control N and P fluxes through river basins, and in the end, to the coastal zone (e.g. Withers & Jarvis, 1998; Hattermann et al., 2006).

Waterlogging, mowing and relative Si uptake

In our experiment, two dominant species (P. arundinacea and A. millefolium, which contain over 1 wt% ASi) and the meadow grass cluster can be identified as Si accumulators. This is consistent with previous literature studies (Hodson et al., 2005). In general, monocots (e.g. Poaceae) contain more Si than dicots. The dicot-monocot division is only a general pattern, and deviations are not an exception. Although a dicot, A. millefolium was previously identified as a substantial accumulator of silicon (Hodson et al., 2005). Interestingly, nearly all the species in the waterlogged treatments had a lower ASi content compared to the non-waterlogged treatments. This contrasts with the Si content in soil solution, displaying higher concentrations in waterlogged anoxic (115 ± 32 μmol l−1) than in oxic conditions (45 ± 5 μmol l−1). Dissolved Si content in soil solution and plant Si content were thus not directly linked in our experiment, although such a link was found in other experiments (e.g. Henriet et al., 2006).

In the experiment, relative uptake of Si in vegetation halved because of mowing practices only. This was mainly because mowing favoured the abundance of the non-Si accumulating species Plantago lanceolata. Si uptake in Si accumulators on the other hand was apparently higher under oxic conditions. This has consequences which might be important to ecosystem engineers, trying to mediate nutrient fluxes. Riparian meadows, established to reduce nutrient fluxes to the coastal zone, might remove significantly less Si, especially relative to N and P, if mowing is applied in spring and if this enhances Si-excluder growth as it did in our experiment. This could be considered a positive effect from a management point of view. Indeed, decreased relative uptake of Si compared to N and P might positively affect N–P–Si ratios in rivers and estuaries during times with risk of DSi depletion.

On the other hand, uptake of Si into grass and/or sedges has been linked to an increased burial of ASi near the surface of floodplain wetlands (Struyf et al., 2009). This increased storage of ASi near the surface can enrich wetland pore-waters with ASi (Struyf et al., 2005, 2009), which can in turn affect the exchange of DSi between wetlands and rivers. The storage of ASi in sediment and litter has been linked to an increased DSi export during low ambient DSi concentrations in tidal rivers (Struyf et al., 2006, 2007). Release of DSi from plant phytoliths has also been shown experimentally (e.g. Fraysse et al., 2010). Reducing the DSi buffering capacity of riparian zones through the yearly removal of the standing crop could potentially lower this Si-buffering capacity of riparian zone to the adjacent rivers. This is an undesired effect in the context of eutrophication.

Perspectives

Our understanding of Si biogeochemistry in riparian meadows, also in those constructed specifically to control nutrient fluxes, is as of yet too limited to evaluate the trade-off between the removal of N and P (and thus also Si) on the one hand, and the creation of potentially Si buffering litter layers (possibly also releasing N and P) on the other hand. In order to implement the control of Si fluxes in the design and management of riparian meadows and small-scale wetlands aimed at reduction of the local output of pollutants and N (e.g. individual factories and households, not directly connected to a general sewer and treatment system), we will need significant progress in our understanding of the functioning of these systems in the Si cycle. Both in situ and ex situ experiments, aimed at testing the importance of specific processes (such as ours) and in situ budget studies, would be valuable in this context.

Interesting topics include:

  • How quick is the recycling of biogenic Si in wetland and meadow soils, and how efficient is this process? Is this recycling complete, or is part of the biogenic Si buried permanently?

  • How much dissolved Si is exchanged between riparian meadows and adjacent rivers, and how tight is the coupling between vegetation and soil biogenic Si?

  • What is the in situ effect of the specific management of soil redox status, vegetation removal, drainage and hydrology?

  • Is it possible to optimize the nutrient ratios released from riparian meadows designed for nutrient removal, by stimulating dissolved Si release to the river, while improving N and P removal?

In the latter context, it is interesting to note that in our experiment, the Si accumulators dominated the waterlogged treatments, resulting in a higher relative Si uptake in vegetation. This will increase the reactive biogenic Si content of the riparian meadow soil, potentially resulting in an increased buffering capacity of Si available in rivers. At the same time, waterlogged conditions might favour nitrogen removal through denitrification (Hill, 1996), but could also favour the release of dissolved orthophosphate into pore-water and thus potentially to the river (Loeb et al., 2008).

Conclusions

Understanding ecosystem resilience and buffer capacity, and how human interference has affected it, is essential to manage negative effects of human impact on these ecosystems (Gunderson, 2000). Land use and vegetation cover have undergone and still undergo huge changes through human intervention. For example, between 50 and 80% of wetlands have been lost from European and North-American estuaries (Elliott & Cutts, 2004), and have been replaced by urban or agricultural surfaces. In Midwestern States (USA), 80% of the wetland acreage has been drained (Mitsch et al., 2001). It has been well documented that this has reduced the resilience of aquatic systems to increased N and P concentrations, increasing the effect of excessive human N and P input. Wetland restoration programs are often based on this understanding.

In our experiments, only a small shift in vegetation coverage, induced by two changed (and manageable) environmental factors, had a relatively strong effect on the Si uptake by plants. This implies that ecosystem managers should take the potential influence of riparian meadows on Si fluxes into account, if these are restored to counteract eutrophication. However, to be able to do so, it is essential to quantify the trade-off between the two effects of Si uptake in riparian vegetation on the Si dynamics in the aquatic environment:
  • Temporary reduction of Si fluxes to rivers through the uptake in silicon accumulating vegetation;

  • The build-up of a biogenic silicon rich surface layer as a result of Si uptake into riparian vegetation, which potentially influences Si dynamics in the adjacent river.

Our concept of Si biogeochemistry in riparian meadows, also for those constructed specifically to control nutrient fluxes, is as of yet too limited to evaluate this trade-off.

Notes

Acknowledgments

L. Clement and E. De Bruyn analyzed Si concentration in samples in the “UA, University of Antwerp, Department of Biology, Testing Laboratory for Chemical Water Quality”. Tom Vanderspiet analyzed N and P concentrations. Eric Struyf (post-doc) and Wout Opdekamp (aspirant) acknowledge personal financing by EU Marie-Curie Programme (FP6) and/or FWO (Research Foundation Flanders). Wiktor Kotowski acknowledges the Francqui Foundation for financing his research work at Antwerp University. We are also thankful to all students, who helped in the species harvesting and sorting.

Open Access

This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.

Supplementary material

10750_2011_737_MOESM1_ESM.doc (23 kb)
Supplementary material 1 (DOC 22 kb)

References

  1. Anderson, D. M., P. M. Glibert & J. M. Burkholder, 2002. Harmful algal blooms and eutrophication: nutrient sources, composition, and consequences. Estuaries 25: 704–726.CrossRefGoogle Scholar
  2. Berner, R. A., 1992. Weathering, plants, and the long-term carbon-cycle. Geochimica et Cosmochimica Acta 56: 3225–3231.CrossRefGoogle Scholar
  3. Blecker, S. W., R. L. McCulley, O. A. Chadwick & E. F. Kelly, 2006. Biologic cycling of silica across a grassland bioclimosequence. Global Biogeochemical Cycles 20: GB3023.Google Scholar
  4. Cloern, J. E., 2001. Our evolving conceptual model of the coastal eutrophication problem. Marine Ecology Progress Series 210: 223–253.CrossRefGoogle Scholar
  5. Conley, D. J., 2002. Terrestrial ecosystems and the global biogeochemical silica cycle. Global Biogeochemical Cycles 16: GB1121.Google Scholar
  6. Conley, D. J., C. L. Schelske & E. F. Stoermer, 1993. Modification of the biogeochemical cycle of silica with eutrophication. Marine Ecology Progress Series 101: 179–192.CrossRefGoogle Scholar
  7. Demaster, D. J., 1981. The supply and accumulation of silica in the marine environment. Geochimica et Cosmochimica Acta 45: 1715–1732.CrossRefGoogle Scholar
  8. Derry, L. A., A. C. Kurtz, K. Ziegler & O. A. Chadwick, 2005. Biological control of terrestrial silica cycling and export fluxes to watersheds. Nature 433: 728–731.PubMedCrossRefGoogle Scholar
  9. Elliott, M. & N. D. Cutts, 2004. Marine habitats: loss and gain, mitigation and compensation. Marine Pollution Bulletin 49: 671–674.PubMedCrossRefGoogle Scholar
  10. Fraysse, F., O. S. Pokrovsky & J. D. Meunier, 2010. Experimental study of terrestrial plant litter interaction with aqueous solutions. Geochimica Et Cosmochimica Acta 74: 70–84.CrossRefGoogle Scholar
  11. Gunderson, L. H., 2000. Ecological resilience – in theory and application. Annual Review of Ecology and Systematics 31: 425–439.CrossRefGoogle Scholar
  12. Güsewell, S., 2005. High nitrogen: phosphorus ratios reduce nutrient retention and second-year growth of wetland sedges. New Phytologist 166: 537–550.PubMedCrossRefGoogle Scholar
  13. Güsewell, S. & M. Koerselman, 2002. Variation in nitrogen and phosphorus concentrations of wetland plants. Perspectives in Plant Ecology, Evolution and Systematics 5: 37–61.CrossRefGoogle Scholar
  14. Hattermann, F. F., V. Krysanova, A. Habeck & A. Bronstert, 2006. Integrating wetlands and riparian zones in river basin modelling. Ecological Modelling 199: 379–392.CrossRefGoogle Scholar
  15. Henriet, C., X. Draye, L. Oppitz, R. Swennen & B. Delvaux, 2006. Effects distribution and uptake of silicon in banana (Musa spp.) under controlled conditions. Plant Soil 287: 359–374.CrossRefGoogle Scholar
  16. Hill, A. R., 1996. Nitrate removal in stream riparian zones. Journal of Environmental Quality 25: 743–755.CrossRefGoogle Scholar
  17. Hodson, M. J., P. J. White, A. Mead & M. R. Broadley, 2005. Phylogenetic variation in the silicon composition of plants. Annals of Botany 96: 1027–1046.PubMedCrossRefGoogle Scholar
  18. Kotowski, W., O. Beauchard, W. Opdekamp, P. Meire & R. van Diggelen, 2010. Waterlogging and canopy interact to control species recruitment in floodplains. Functional Ecology 24(4): 918–926.CrossRefGoogle Scholar
  19. Loeb, R., L. P. M. Lamers & J. G. M. Roelofs, 2008. Effects of winter versus summer flooding and subsequent desiccation on soil chemistry in a riverine hay meadow. Geoderma 145: 84–90.CrossRefGoogle Scholar
  20. Lotze, H. K., H. S. Lenihan, B. J. Bourque, R. H. Bradbury, R. G. Cooke, M. C. Kay, S. M. Kidwell, M. X. Kirby, C. K. Peterson & J. B. C. Jackson, 2006. Depletion, degradation, and recovery potential of estuaries and coastal seas. Science 312: 1806–1809.PubMedCrossRefGoogle Scholar
  21. Mitsch, W. J., J. W. Day, J. W. Gilliam, P. M. Groffman, D. L. Hey, G. W. Randall & N. Wang, 2001. Reducing nitrogen loading to the gulf of Mexico from the Mississippi river basin: strategies to counter a persistent ecological problem. Bioscience 51: 373–388.CrossRefGoogle Scholar
  22. Olde Venterink, H., J. E. Vermaat, M. Pronk, F. Wiegman, G. E. M. van der Lee, M. W. van den Hoorn, L. Higler & J. T. A. Verhoeven, 2006. Importance of sediment deposition and denitrification for nutrient retention in floodplain wetlands. Applied Vegetation Science 9: 163–174.CrossRefGoogle Scholar
  23. Sommer, M., D. Kaczorek, Y. Kuzyakov & J. Breuer, 2006. Silicon pools and fluxes in soils and landscapes – a review. Journal of Plant Nutrition and Soil Science – Zeitschrift Fur Pflanzenernahrung Und Bodenkunde 169: 582.CrossRefGoogle Scholar
  24. SPSS, 2005. SPSS version 14.0. SPSS, Chicago, Illinois, USA.Google Scholar
  25. Street-Perrott, A. F. & P. A. Barker, 2008. Biogenic silica: a neglected component of the coupled global continental biogeochemical cycles of carbon and silicon. Earth Surface Processes and Landforms 33: 1436–1457.CrossRefGoogle Scholar
  26. Struyf, E. & D. J. Conley, 2009. Silica: an essential nutrient in wetland biogeochemistry. Frontiers in Ecology and Environment 7: 88–94.CrossRefGoogle Scholar
  27. Struyf, E., S. Van Damme, B. Gribsholt, J. J. Middelburg & P. Meire, 2005. Biogenic silica in tidal freshwater marsh sediments and vegetation (Schelde estuary, Belgium). Marine Ecology Progress Series 303: 51–60.CrossRefGoogle Scholar
  28. Struyf, E., A. Dausse, S. Van Damme, K. Bal, B. Gribsholt, H. T. S. Boschker, J. J. Middelburg & P. Meire, 2006. Tidal marshes and biogenic silica recycling at the land-sea interface. Limnology & Oceanography 51: 838–846.CrossRefGoogle Scholar
  29. Struyf, E., S. Van Damme, B. Gribsholt, K. Bal, O. Beauchard, J. J. Middelburg & P. Meire, 2007. Phragmites australis and Si cycling in tidal wetlands. Aquatic Botany 87: 134–140.CrossRefGoogle Scholar
  30. Struyf, E., W. Opdekamp, H. Backx, S. Jacobs, D. J. Conley & P. Meire, 2009. Vegetation and proximity to the river control amorphous Si storage in a riparian wetland (Bierbza National Park, Poland). Biogeosciences 6: 623–631.CrossRefGoogle Scholar
  31. Tréguer, P. & P. Pondaven, 2000. Global change – silica control of carbon dioxide. Nature 406: 358–359.PubMedCrossRefGoogle Scholar
  32. Walinga, I., W. Van Vark, V. J. G. Houba & J. J. Van Der Lee, 1989. Plant Analysis Procedures. Soil and Plant Analysis, Part 7, pp. 13–16. Agricultural University, Wageningen, NL: 13–16.Google Scholar
  33. Withers, P. J. A. & S. C. Jarvis, 1998. Mitigation options for diffuse phosphorus loss to water. Soil Use and Management 14: 186–192.CrossRefGoogle Scholar

Copyright information

© The Author(s) 2011

Authors and Affiliations

  • Eric Struyf
    • 1
  • Wiktor Kotowski
    • 2
  • Sander Jacobs
    • 1
  • Stefan Van Damme
    • 1
  • Kris Bal
    • 1
  • Wout Opdekamp
    • 1
  • Hans Backx
    • 1
  • Dimitri Van Pelt
    • 1
  • Patrick Meire
    • 1
  1. 1.Department of Biology, Ecosystem Management Research GroupUniversity of AntwerpAntwerpBelgium
  2. 2.Department of Plant Ecology and Environmental ConservationUniversity of WarsawWarsawPoland

Personalised recommendations