JEQ Grow Your Career With ASA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 7 November 2005
Published in J Environ Qual 34:2145-2155 (2005)
DOI: 10.2134/jeq2004.0466
© 2005 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Braskerud, B. C.
Right arrow Articles by Koskiaho, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Braskerud, B. C.
Right arrow Articles by Koskiaho, J.
Agricola
Right arrow Articles by Braskerud, B. C.
Right arrow Articles by Koskiaho, J.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Ecosystem Restoration
Right arrow Best Management Practices
Right arrow Phosphorus
Right arrow Water Pollution

Phosphorus Workshop

Can Constructed Wetlands Reduce the Diffuse Phosphorus Loads to Eutrophic Water in Cold Temperate Regions?

B. C. Braskeruda,*, K. S. Tonderskib, B. Weddingc, R. Bakked, A.-G. B. Blankenberga, B. Uléne and J. Koskiahof

a Jordforsk, Frederik A. Dahls vei 20, NO-1432 Ås, Norway
b Linköping University, IFM/Biology, SE-581 83 Linköping, Sweden
c Ekologgruppen i Landskrona AB, Järnvägsgatan 19B, SE-261 32 Landskrona, Sweden
d Telemark University College, Box 203, NO-3901 Porsgrunn, Norway
e Division of Water Quality Management, Swedish University of Agricultural Sciences, Box 7072, SE-750 07 Uppsala, Sweden
f Finnish Environment Institute, Integrated River Basin Research, P.O. Box 140, FIN-00251 Helsinki, Finland

* Corresponding author (bcb{at}nve.no)

Received for publication December 8, 2004.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Construction of wetlands is a possible supplement to best management practices (BMP) at the field level to mitigate phosphorus (P) pollution from agricultural areas. In this paper, annual results from 17 intensively studied wetlands in the cold temperate or boreal climatic zone are reported and analyzed. Surface areas varied from 0.007 to 8.7% of the catchment area. The average total phosphorus (TP) retention varied from 1 to 88%, and the dissolved reactive phosphorus (DRP) retention from –19 to 89%. Retention varied substantially from site to site, indicating the existence of site-specific factors in the catchment and wetlands that influenced the P removal. Factors important for P retention in wetlands were evaluated through multiple statistical analyses by dividing P into two fractions: particulate phosphorus (PP) and DRP. Both relative (%) PP and DRP retention increased with wetland surface area. However, PP retention was not as sensitive as DRP in terms of wetland size and retention: specific PP retention (gram P retention per m2 and year) decreased as wetland area (Aw) increased, suggesting the existence of a site-specific optimal wetland to catchment area (Ac) ratio. Particulate P retention decreased with increasing DRP to TP ratio, while the opposite was found for DRP. Dissolved reactive P retention was higher in new than in old wetlands, while increasing age did not influence PP retention negatively. Effective BMP in the catchment is important to keep the P loss low, because the outlet concentration of P from wetlands is often positively correlated to the input concentration. However, wetlands act as the last buffer in a catchment, since the retention often increases as the P concentration in streams increases.

Abbreviations: Ac, catchment area • Aw, wetland area • DRP, dissolved reactive phosphorus • HLR, hydraulic loading rate • k, (first-order) removal rate constant • PCA, principal component analysis • PP, particulate phosphorus • q, specific runoff from catchment • TP, total phosphorus • w', phosphorus settling velocity


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
MEASURES TO REDUCE LOSSES of P from agricultural areas may be limited and have been poorly investigated (Withers and Lord, 2002). In addition, background loads of P from arable land may be relatively high (Ulén et al., 2005). The effects of phosphorus in recipient waters depend on whether P is associated to particles (PP) or dissolved. Dissolved reactive P can be utilized more easily by algae than PP and is therefore of greater significance in the immediate triggering of algal blooms. However, Uusitalo et al. (2003) showed that PP lost from clay soils through surface runoff and drainage water made a significant contribution to bioavailable P. Approximately 6 to 10% of the PP was available for immediate use to algae. Since the PP fraction of TP was large, the algae-available P in PP was as important as DRP in the discharge water. In a similar way, Krogstad and Løvstad (1989) showed that 25 to 75% of PP was accessible for algae. However, PP is deposited in sediments to a larger extent and hence is usually less readily available to algae, although large algal blooms trigger the release of P from sediments through biological and chemical reactions, enhancing the blooms in an escalating manner (e.g., Knuuttila et al., 1994). The total available P as the sum of DRP in the water column and P released from sediments determines the extent and severity of the blooms. Uusitalo et al. (2003) showed that 34 to 55% of PP was solubilized through changes in redox conditions in the sediments. Reduction of both total P and DRP loads on eutrophic lakes is therefore significant in attempts to reduce the magnitude and the negative effects of algal blooms (Correll, 1999).

A recent review showed that free water surface wetlands (FWS) can remove phosphorus loading and provide low-cost phosphorus removal (Reddy et al., 1999). However, a large variation in the area-specific (g P retention m–2 yr–1) or relative (g P retention per g P load) TP retention was found in this review. In Florida, much effort has been put into studies of shallow wetlands dominated by submerged macrophytes, and quite promising results have been obtained (e.g., Dierberg et al., 2002). It is clear that macrophyte-mediated pH increases and subsequent precipitation of calcium carbonate with co-precipitation of phosphates play a dominant role in the overall P removal in those wetlands. This mechanism plays a less significant role in P removal in temperate wetlands with extended cold periods, where photosynthetic increases in pH only occur during a few months, and emergent vegetation dominates. However, small and shallow constructed wetlands in Norway have proven to act as efficient P traps in predominantly agricultural catchments (Braskerud, 2002a), implying that wetland P removal is also possible in temperate regions. These wetlands had a high specific TP removal (26–71 g m–2 yr–1) and high hydraulic loads (0.7–1.8 m d–1). However, the content of suspended particles in incoming water was fairly high, suggesting that net removal was mainly related to retention of particles by sedimentation. In contrast, Jordan et al. (2003) reported no net removal of total P and suspended solids in a hydraulically low-loaded wetland in the United States. A literature study of 49 natural wetlands world wide showed that 41 had a net relative TP retention of 58% (average; SD = 23%), while 3 had no effect and 5 had a net release (Fisher and Acreman, 2004).

An obviously important factor is the form in which the P enters the wetland. Wetlands act as efficient traps for particles (Braskerud, 2001) and P adhering to particles is efficiently captured in the short-term perspective. However, it is less well documented whether wetlands also retain dissolved P, although Carleton et al. (2001) reported net removal of ortho-phosphate for the majority of the 16 wetland data sets they evaluated. Promising results on removal of reactive P have also been reported by Liikanen et al. (2004). Fisher and Acreman (2004) divided their natural wetlands into swamps and marshes and riparian wetlands. Most studies of swamps and marshes showed a net retention of soluble P, while a loss was often recorded for riparian wetlands.

In the present study, our objectives were to:


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Description of Wetlands
Seventeen constructed wetlands receiving agricultural runoff were selected, constituting a set of data covering a large span of constructed wetland surface areas, loads, and ratios of wetland surface area to catchment area (Table 1). Catchments varied from 0.05 to 8.8 km2, and the size of the wetlands varied from 347 to 10000 m2. The resulting ratio of wetland to catchment size varied from 0.007 to 8.7%.


View this table:
[in this window]
[in a new window]
  		Table 1. Characteristics of 17 constructed wetlands included in this study, sorted according to ratio of surface area to catchment area.

 
All wetlands examined were free water surface systems (FWS) (i.e., systems in which water ran through a pond-like structure), except Site 7, which also included some filters. The vegetation cover of the different wetlands varied, and did not cover the whole sediment area in wetlands with less than 4 yr of sampling.

The original study by Kovacic et al. (2000) included three wetlands, which were constructed 1 yr before sampling started. The last wetland (D) had a negative TP retention of –26% on average. The negative retention in this wetland was questionable, since the TP export was larger than the input of phosphorus in two out of three years. Kovacic et al. (2000) suggested a possible error in the particulate P flux registration, so we omitted this wetland from our dataset.

Sampling Systems
At all experimental sites runoff was monitored continuously. The sampling programs were intensive, which is needed in small watersheds with dynamic runoff. Usually volume proportional composite samples were taken, but at some sites the data represent time integrated composite samples. Grab samples were used temporarily when frost impeded the automatic sampling (Sites 11 and 15), or when the automatic sampling failed (Sites 5, 8, and 9), and permanently in the outlet of relative large wetlands (Sites 14 and 16).

Water Analysis
National standard methods were used for the analyses. The water was filtered with different filters before DRP was analyzed. In most cases, filters with a pore size of 0.45 µm were used (Table 1), but on some occasions filters with finer pores (Site 1, 0.2 µm) or coarser pores (Site 12, 0.75 µm) were also used. In three Swedish wetlands (Sites 5, 8, and 9), ortho-P was measured on unfiltered samples. Only data on filtered samples were used in the statistical analyses of DRP.

Two possible errors may have affected the results on the DRP retention:

Despite this, automatic sampling systems are the only way to make reliable mass balance budgets for wetlands located in small catchments, since the P flux is very dynamic (e.g., Braskerud, 2002a; Wedding, 2003). As an example, Johnson (1992) calculated a need for 600 to 1000 spot samples over a 3-yr period to characterize sediment transport from small catchments. In addition, the proportion of DRP vs. TP in surface water may vary rather widely between sites, and although the absolute values of DRP may be erroneous, the relative difference between catchments may still be reflected in data from automatic sampling programs. Thus, there was sufficient motivation for analyzing the possible importance of the DRP fraction on differences in TP retention between wetlands.

Estimation of the Removal Rate Constant (k) or Phosphorus Settling Velocity (w')
The first-order area model is widely used to describe P removal in wetlands. According to Kadlec and Knight (1996), the first-order area model is:

[1]
if the background concentration (i.e., the concentration resulting from wetland internal transformation processes) is neglected. Here k is the removal rate constant (m yr–1), HLR is the hydraulic loading rate (m yr–1), and Cin and Cout are the concentrations of P in inlet and outlet water (mg L–1). The HLR is the annual runoff divided by wetland surface area (Aw). For settling of particles, the removal rate constant k is similar to the settling velocity w (Kadlec and Knight, 1996; Moustafa et al., 1996; Braskerud, 2002b). In catchments where a large part of the TP is PP, k can be used as an approximation of the P settling rate w'.

Statistical Analyses
Simple linear regression and principal component analysis (PCA) and factor analyses were applied (JMP 3.1.5; SAS Institute, 1989) using the annual values from all wetlands. Principal component analysis and factor analyses are used to seek the underlying explanations that simple linear regressions are unable to provide. It is the combination of the variables within the individual factors that are important: the parameters included in each factor, how they influence each other, and the magnitude of the factor loadings. A large positive or negative factor loading indicates a significant influence of the variable. However, factor loadings cannot be split into statistically significant or not. The PCA analyses were performed on correlations to standardize the variables. Rotated factor analyses were conducted on factors with an Eigen value larger than 1 (here: four factors). For reasons of clarity, the + and – signs in Factor 3 (Table 4) were reversed.


View this table:
[in this window]
[in a new window]
  		Table 4. Rotated principal component analysis (PCA) and factor analyses on correlations, showing the most important variables affecting the phosphorus retention in 16 ponds and wetlands (n = 73 yr). Site 12 was excluded. Variables that were most important (<–0.6 or >0.6) are in italic type, while less important variables have been deleted (–0.25 to 0.25).

 

   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Average Retention of Total Phosphorus and Dissolved Reactive Phosphorus
The specific TP load to the wetlands varied between 0.7 and 307 g m–2 yr–1, with the highest load at Site 1. However, this wetland received low average TP concentrations (Table 2) and a very high hydraulic load. The average P concentration varied substantially between the sites studied (0.07–2.15 mg L–1), with a mean value of 0.4 mg L–1 (median 0.2 mg L–1). In contrast to Site 1, some of the wetlands receiving small loads (e.g., Site 17) received water with very high concentrations of TP and extremely low hydraulic loads. Site 12 probably received some wastewater, while Site 17 received water from an intensively cropped potato field. Notably, even though the TP retention could be negative in some years, all wetlands showed a net retention of TP. The variation between wetlands in the proportion of DRP in inflow was large, with a mean of 51% and a median of 38%, suggesting considerable differences in catchment characteristics.


View this table:
[in this window]
[in a new window]
  		Table 2. Average hydraulic loading rate (HLR), first-order rate constant (k), phosphorus (P) concentration (±standard deviation), and specific P load for 17 constructed wetlands. Relative and specific phosphorus retention are presented, as are the ratio DRP x 100 TP–1 in stream water and retention of DRP (where data are available), where TP is total phosphorus and DRP is dissolved reactive phosphorus.

 
Retention of DRP was observed in most cases, but was highly variable, ranging from –19 to 89%. Similarly high retention (82%) was also observed in several campaign measurements at Site 17 (Higgins et al., 1993). Site 12 showed a remarkably high DRP retention (89%). As a result, Site 12 often acted as an outlier in the statistical analyses and was excluded from the statistical analyses below. Data on unfiltered ortho-P retention were only available for the Sites 5, 8, and 9 (8–37% total reactive P retention).

Factors Affecting the Phosphorus Retention
A more detailed analysis of the factors affecting the TP retention was made using PCA and factor analyses. Retention was expressed in three ways:

The PCA and factor analyses were based on a correlation matrix (r) (Table 3). Note that most of the variables in the r matrix are not normally distributed. However, it is still possible to carry out PCA and factor analyses (e.g., Johnson and Wichern, 1992) (Table 4). Tables 3 and 4 include annual data from all the wetlands in Table 2 except Site 12.


View this table:
[in this window]
[in a new window]
  		Table 3. Correlation matrix (r) for variables affecting the total phosphorus (TP) retention in 16 wetlands (Site 12 excluded) (n = 73 yr).{dagger}

 
Each factor (1–4) in Table 4 shows the internal dependency between the variables on a scale from –1 to 1. Within one factor, factor loads with the same sign are positively correlated while an opposing sign indicates a negative correlation. Factor loads are the numbers in the columns (Factors 1–4). Communalities show the part of the variable described by Factors 1–4 totally. Factor loads larger than 0.6 or less than –0.6 are presented in italic type, while factor loads in the interval –0.25 to 0.25 were deleted to highlight the most important variables (Table 4).

The PCA and factor analyses described approximately 80% of the variation, where the influence of the factors decreased in the order Factor 1 > Factor 2 > Factor 3 > Factor 4. The influence of the individual variables within the factors was:

Factors 1 and 2 from the PCA and factor analyses are presented graphically in Fig. 1 . The figure was composed by setting the factor loadings in Factor 1 and Factor 2 as (x,y) coordinates, respectively.



View larger version (11K):
[in this window]
[in a new window]
  		Fig. 1. Plot of the results from the two first factors in the principal component analysis (PCA) and factor analyses. TP%, relative total phosphorus retention; TPSL, specific total phosphorus load; TPSR, specific total phosphorus retention; TPCI, inlet total phosphorus concentration; TPCO, outlet total phosphorus concentration; q, specific runoff from catchment; HLR, hydraulic loading rate; DRP, dissolved reactive phosphorus; Aw to Ac ratio, wetland surface area to catchment area ratio; age, years from construction of wetland; k, first-order removal rate constant.

 
Retention of Total Phosphorus and Dissolved Reactive Phosphorus
In this section, some results on TP and DRP retention are highlighted. Ideally, it is possible to estimate PP by subtracting DRP from TP. However, errors introduced by the sampling program design tend to increase when this is done. In addition, DRP data were lacking for several wetlands (see Table 2). As a result, we chose to highlight the more reliable TP data but some results from the data on DRP were also included to help us identify important relationships in wetland P retention performance and the influence of catchment characteristics on this. The absolute values in the DRP results should be used with care, though we argue that they reflect true differences between wetlands with respect to DRP as a proportion of total P. Note that (i) DRP data were only available from the wetlands with DRP retention in Table 2 (n = 46 yr), and (ii) the same scale is not used in all diagrams. For Site 13 the DRP sampling was close to the ISO 5667 standard (International Organization for Standardization, 1994), while most of the P was DRP at Sites 14 and 16. Hence, the results from these wetlands were probably not influenced by the errors indicated in the Materials and Methods, and are highlighted in some of the DRP figures.

Catchment Influence on Phosphorus Retention
The average input concentration of TP varied over a range of three orders of magnitude between the catchment streams (Table 2). Increased TP concentration in the wetland inlets increased the relative retention of TP (Fig. 2a) . The same was true for an increase in the DRP concentration. Note that each data point represents the mean value of 1 yr in all figures.



View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2. The relationship between wetland relative phosphorus retention and concentration in the inlet of (a) total phosphorus (TP) and (b) dissolved reactive phosphorus (DRP). Results with minor errors in DRP are circled.

 
The DRP to TP ratio varied considerably between the catchments (Table 2), and gives an inverse indication of the proportion of the P that was sorbed to particles and could undergo sedimentation (Fig. 3) : TP retention decreased as the dissolved part of TP in the inlet increased (Fig. 3a), while the DRP retention increased (Fig. 3b). The first result is observed in Factor 1 and 4 (specific retention and outlet concentration; Table 4). The PCA and factor analyses produced no factors where the P retention increased as the DRP to TP ratio increased, with one possible exception: the concentration in the outlet decreased as the DRP to TP ratio increased (i.e., as the PP fraction decreased).



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 3. Relative retention of (a) total phosphorus (TP) and (b) dissolved reactive phosphorus (DRP) as a function of DRP to TP ratio in the inlet. Results with minor errors in DRP are circled. Site 12 was not included in the statistics (+ in box).

 
The removal rate constant k indicates how efficient the P retention performance is (see Eq. [1]). The constant is estimated from the TP data, and thus includes all types of P retention processes. Figure 4 shows that k decreased as the DRP to TP ratio increased (i.e., as the PP fraction in the inlet water decreased). As a result, k was positively related to the specific TP retention, whereas there was no correlation with the specific DRP retention (Fig. 5) .



View larger version (20K):
[in this window]
[in a new window]
  		Fig. 4. Relationship between the first-order rate constant (k) value and dissolved reactive phosphorus (DRP) to total phosphorus (TP) ratio. Site 12 was not included in the statistics (+ in box).

 


View larger version (15K):
[in this window]
[in a new window]
  		Fig. 5. The relationship between the first-order rate constant (k) and (a) specific total phosphorus (TP) retention and (b) dissolved reactive phosphorus (DRP) retention.

 
Wetland Influence on Phosphorus Retention
As the ratio of wetland surface area (Aw) to catchment area increased, the relative P retention increased, as expected (Fig. 6) . The specific TP retention, however, followed a different pattern (Fig. 7a) : for wetlands with a size larger than roughly 0.04% of the catchment area, the specific retention decreased with an increase in size. Data from the small wetland at Site 1, however, suggest that there is minimum size below which the specific retention will drop rapidly. The specific DRP retention had no statistically significant correlation with changes in the wetland surface area (Fig. 7b).



View larger version (16K):
[in this window]
[in a new window]
  		Fig. 6. The relative retention of (a) total phosphorus (TP) and (b) dissolved reactive phosphorus (DRP) as a function of wetland area (Aw) to catchment area (Ac) ratio. Results with minor errors in DRP are circled. Site 12 was excluded from the statistical analyses (+ in square).

 


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 7. Relationship between specific retention of (a) total phosphorus (TP) and (b) dissolved reactive phosphorus (DRP) and wetland area (Aw) to catchment area (Ac) ratio. Dotted curve represents an estimate. Site 12 was included neither in the statistics, nor in (b), since the result was in the same order of magnitude as the TP retention.

 
There has been a concern among wetland scientists that the P binding capacity in the sediments will be saturated, converting the wetlands to P sources rather than P sinks. The correlation matrix (Table 3) indicated no significant correlation between age and TP retention as seen in Fig. 8a . In contrast, the PCA analyses indicated a positive effect of age in Factor 1, and the opposite in Factor 2 (Table 4). A negative effect of age was observed for DRP in Fig. 8b.



View larger version (18K):
[in this window]
[in a new window]
  		Fig. 8. Retention of (a) total phosphorus (TP) and (b) dissolved reactive phosphorus (DRP) as a function of wetland age. Sites with only 1 yr of observations were excluded. Results with minor errors in DRP are circled.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Use of wetlands as a best management practice (BMP) tool to combat diffuse pollution from agricultural catchments should include evaluation of the effects on both PP and DRP. The use of sampling procedures according to international standards (ISO 5667; International Organization for Standardization, 1994) may lead to methodological problems regarding DRP analysis, since the equilibrium between PP and DRP may change due to the long time between sampling and filtration when using composite samples. As a result DRP data probably either do not represent P flux for mass balance studies due to too few sampling occasions, or the equilibrium between PP and DRP may have changed during storage in the composite sampling container. Therefore, in the present study the main focus was on TP in the statistical analyses, but the data on DRP were used to indicate differences between wetlands in the proportion of DRP and how that related to differences in overall TP retention. Regarding the importance of the proportion of DRP in the water, we acknowledge that the data we used were probably subject to some errors, as discussed above. Furthermore, we did not have any basis for discussing the difference between truly dissolved P and P adsorbed to fine particles like colloids that may also not settle in treatment wetlands. They were therefore included in our discussion of the importance of DRP.

In addition the results were based on annual observations. Studying the P retention with a higher time resolution (i.e., seasonal variation) would add to the understanding of the processes going on in each wetland.

Dividing Total Phosphorus into Particulate Phosphorus and Dissolved Reactive Phosphorus
There was generally a high correlation between the results presented in the r matrix (Table 3) and the PCA and factor analyses (Table 4). However, the PCA and factor analyses did change the influence of some variables on the relative and specific TP retention, for example, wetland age, Aw to Ac ratio, DRP to TP ratio, and k. This could be due to the complex nature of phosphorus species and to the variety of P retention processes in the wetlands (e.g., adsorption to particles, sedimentation, uptake in biota, etc.).

It is possible that the PCA and factor analyses grouped the retention systems for different P species within the four factors. Factor 1 (Table 4) could highlight PP retention, while Factor 2 includes more of the DRP retention. These two factors are highlighted in Fig. 1. We had no data to group other P fractions. Our justification for grouping Factors 1 and 2 into PP and DRP retention, respectively, is:

Overall, this suggests that Factor 1 in the PCA and factor analyses mainly represented removal of PP, whereas Factor 2 also distinguished factors important for removal of other fractions of P. These other fractions of P may include small sized PP, but probably not aggregates or particles in the silt and sand fraction, because these fractions easily settle in even small wetlands (Braskerud, 2003). Dividing Factors 1 and 2 into mainly PP and DRP retention, respectively, makes it possible to distinguish between processes important for PP and DRP retention.

Effect of Concentration and Loading Rate
The specific TP retention was influenced by the TP load (Factor 1, Table 4 and Fig. 1). Several authors have found a positive relationship between load and retention (e.g., Moustafa et al., 1996; Kadlec and Knight, 1996; Braskerud, 2002a). However, our analyses suggested that variables other than the load were more important for the specific retention.

The relative DRP retention increased with increased DRP concentration (Fig. 2b). This result is in contrast to the findings at Site 15, where the relative DRP retention decreased as the DRP concentration increased (Liikanen et al., 2004). The latter explained their results by a relative reduction in the chemical sorption reactions to Al and Fe oxides as the P content in water increased. In our dataset, however, increased TP concentration increased the relative TP retention (Factor 2 and Fig. 2a). An increase in the DRP concentration could lead to a relatively higher possibility of P being sorbed to particles or biota (e.g., phytoplankton and floating microalgae at Site 13; Reinhardt et al., 2005).

Effect of Surface Area
In the present study, the relative TP retention increased with larger ratio of wetland area to catchment area (Aw to Ac) (Factor 2, Table 4; Fig. 1; Fig. 6a), whereas the specific TP retention decreased as a result of the lower specific load on the larger wetlands (Factor 1, Table 4; Fig. 1; Fig. 7a). This raises an important question: What is the optimal size from a cost efficiency point of view? In very small wetlands like Site 1, resuspension of sediments is likely to occur during storm runoff events. It is likely that the specific retention reaches a maximum at an Aw to Ac ratio between 0.01 and 0.06% and then decreases as the ratio increases (Fig. 5b). Negative retention has been observed in some large constructed (like Site 14) and natural wetlands (Fisher and Acreman, 2004). According to Fig. 6a, the relative TP retention increased rapidly up to an Aw to Ac ratio of 1 to 2%, before it leveled out.

The wetland surface area may be less important for PP retention than for DRP retention. The PP retention decreased as the Aw to Ac ratio increased according to Factor 1. The specific TP retention associated with this factor is probably the main reason for this result. Table 1 and 2 show that a wetland with an Aw to Ac ratio of 0.06% (Site 3) performed as well as a wetland with an Aw to Ac ratio seven times higher (Site 10). Hence, catchments dominated by PP runoff may use smaller wetlands than catchments dominated by DRP losses.

When the Aw to Ac ratio increased, the relative DRP retention had a tendency to increase in a similar way as for TP (Fig. 6b). For the specific DRP retention, however, there was no statistically significant fit with Aw to Ac ratio (Fig. 7b). Our data do not indicate an upper limit of the Aw to Ac ratio for DRP retention. Reinhardt et al. (2005) estimated an Aw to Ac ratio of 4% to retain half of the DRP input. This corresponds with a detention time of approximately 7 to 10 d.

Effect of Dissolved Reactive Phosphorus to Total Phosphorus Ratio
A significant decrease in TP retention was observed as the ratio of DRP to TP increased (Factor 1 and Fig. 3a), showing the importance of inflow quality on wetland TP retention capacity. This is reasonable if we assume that sedimentation of PP is the dominant process for TP retention (see Fig. 4).

For DRP, the results showed a positive relationship between relative retention and the DRP to TP ratio (Fig. 3b). This is somewhat controversial, since we normally expect decreased DRP retention as the ratio of DRP to TP increases, because less particles are available for P sorption. These observations can be explained as increased uptake by biota, since a larger part of TP is available. Braskerud (2002a) found more TP retention during summer than during other seasons at Sites 3, 4, 6, and 10, which could support this assumption. Reinhardt et al. (2005) also included uptake by biota as a major DRP retention process (Site 13). At Sites 3, 4, 6, and 10, the effect of season was small compared to other variables, while it was very large at Site 13. Due to the short summer in the cold temperate zone, it is likely that uptake by biota usually plays a minor role compared to other DRP retention processes, such as adsorption to particles and sedimentation. In addition, most of the P enters the wetlands in the autumn storms or when snow melts in the spring. In these periods, biological activity is regarded as insignificant.

The interpretation of Fig. 3b is complex due to the possible errors in the DRP content arising from delayed filtration after sampling and clogging of filters (see Materials and Methods). However, some of the results indicate a substantial DRP retention:

These hypotheses call for further research, as they are of significance for the effect and use of such treatment systems.

Effect of Age
The PCA and factor analyses suggested that age had a positive influence on the specific TP retention (Factor 1, which we interpreted as PP retention). This might be an indirect effect of increased vegetation cover, since most of the wetlands were dominated by emergent vegetation. Braskerud (2001) found that resuspension of the wetland sediment decreased as vegetation cover increased, and that the positive effect of vegetation on retention increased until the vegetation cover was approximately 50% of the surface area. The simple regression analysis did not show any influence of age on the relative TP retention (Fig. 8a).

In contrast to the TP retention, the relative DRP retention decreased with wetland age (Factor 2 and Fig. 8b). However, none of the individual sites had a statistically significant reduction in DRP retention over time (data not shown). Since the DRP results were biased, we could not predict a cessation of DRP retention after 9 yr (Fig. 8b). However, Factor 2 (which we interpreted as DRP retention) indicates a possible problem not fully understood. The concern that wetlands may become sources of P calls for more long-term investigations.

The Importance of Runoff and Hydraulic Load
In the present PCA and factor analyses, only the PP retention (Factor 1) was influenced by the specific runoff (q), but not the hydraulic loading rate (HLR). Note that the communalities of q were rather low. For DRP (Factor 2) the retention was related neither to q nor HLR. Similarly, Carleton et al. (2001) found no relationship between TP removal and HLR, and concluded that TP removal was more related to nominal detention time and to the Aw to Ac ratio. Our data support this result. However, if the time resolution was smaller, HLR would probably be a better variable than the Aw to Ac ratio (e.g., Braskerud, 2002a; Reinhardt et al., 2005).

Runoff is important for the redox conditions in free surface water wetlands. A study at Site 3 showed that the redox potential in the wetland outlet water was always positive and often rather high (median 550 mV; Braskerud et al., 2005). High hydraulic loads supplied the wetland with sufficient water to keep it aerobic. As a result, periods with P loss from the wetland were rare, because oxygen-rich water creates a protecting oxidized microlayer above the sediments (Penn et al., 2000; Braskerud et al., 2005).

In general, the relative retention increases as the Aw to Ac ratio increases (Fig. 6). Increased surface area could, however, reduce the redox potential since HLR decreases. This increases the likelihood of P being desorbed from the deeper sediments, even though convective circulation of water due to diurnal temperature variations between surface water and near-bottom water may occur under stagnant periods, as shown at Site 15 (Schmid et al., 2005). As a result, wetlands would occasionally show P leakage from the sediment. However, retention of P rich soil particles in wetlands with high HLR may be relatively permanent.

The Impact of Site
Catchments are often unique with respect to runoff, P loss, and DRP to TP ratio, as we have observed in this study (Tables 1 and 2), and in the literature (e.g., McDowell et al., 2001). The P retention varied considerably between catchments with wetlands, despite similar Aw to Ac ratios (Fig. 5 and 6).

Site 12 was treated as an outlier in the statistical analyses since the P load, P concentration, and DRP to TP ratio were unusually high. Nevertheless, this wetland system had a remarkably good retention performance (Table 2). This was probably due to the mineral phyllite in the soil of the catchment. This is likely to have given the stream a high content of iron, which would facilitate P retention in the wetland system by forming P–Fe complexes.

Braskerud (2003) showed that clay particle retention was highly dependent on human activity in the catchments. A clear example is Site 4, where the agricultural land had been artificially leveled (i.e., the soil had been roughly scraped to deep layers). Due to aggregate destruction, the subsequent wetland had a significantly lower (22%) clay particle retention than Site 3 (57%), which was less disturbed. A comparison revealed that the relative retention of P was lowest in the leveled catchment (Table 2).

Less Phosphorus to Receiving Water
Unfortunately, a reduction in P load to lakes is often not enough to improve the water quality, since much P is already stored in the lake sediment and more may be released under certain conditions than the annual external input (e.g., Knuuttila et al., 1994; Moore et al., 1998). However, a load reduction is necessary for the long-term goal of cleaner water. Factor 4 (Table 4) shows that the P concentration in the wetland outlet follows the inlet concentration. Hence, the best way to improve the water quality is to reduce the P losses from agricultural areas. However, even areas that fulfil the best management practice goals will leach P. In this respect constructed wetlands may be an important part of the measures taken in a catchment.

Catchments are unique with respect to P losses, so we expect that the retention performance of wetlands will also be unique. This must be borne in mind when constructing wetlands to combat diffuse P pollution. For some catchments construction of wetlands may be cost-effective, for others not.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The study of 17 constructed ponds and wetlands in the cold temperate and boreal climatic zones showed that:


   ACKNOWLEDGMENTS
 
We would like to thank the Norwegian Agricultural Authority, the Norwegian Research Council, and the Foundation for Strategic Environmental Research through the Swedish Water Management Research Programme (VASTRA) for their financial support. We also appreciate the constructive comments from our reviewers, and Mary McAfee for helping to improve the English.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 




This article has been cited by other articles:


Home page
 J. Environ. Qual.Home page
C. C. Hoffmann, C. Kjaergaard, J. Uusi-Kamppa, H. C. B. Hansen, and B. Kronvang
Phosphorus Retention in Riparian Buffers: Review of Their Efficiency
J. Environ. Qual., August 24, 2009; 38(5): 1942 - 1955.
[Abstract] [Full Text] [PDF]

Home page
 J. Environ. Qual.Home page
Y. Chen, R. C. Viadero Jr., X. Wei, R. Fortney, L. B. Hedrick, S. A. Welsh, J. T. Anderson, and L.-S. Lin
Effects of Highway Construction on Stream Water Quality and Macroinvertebrate Condition in a Mid-Atlantic Highlands Watershed, USA
J. Environ. Qual., June 23, 2009; 38(4): 1672 - 1682.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (7)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Braskerud, B. C.
Right arrow Articles by Koskiaho, J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Braskerud, B. C.
Right arrow Articles by Koskiaho, J.
Agricola
Right arrow Articles by Braskerud, B. C.
Right arrow Articles by Koskiaho, J.
Related Collections
Right arrow Wetlands and Aquatic Processes
Right arrow Ecosystem Restoration
Right arrow Best Management Practices
Right arrow Phosphorus
Right arrow Water Pollution

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Soil Science Society of America Journal Journal of Plant Registrations The Plant Genome