| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Wetland Biogeochemistry Lab., Soil and Water Science Dep., Univ. of Florida; 106 Newell Hall, P.O. Box 110510, Gainesville, FL 32611
b Dep. of Oceanography and Coastal Sciences, Wetland and Aquatic Biogeochemistry Lab., Louisiana State Univ., Baton Rouge, LA 70803
* Corresponding author (jrwhite{at}lsu.edu).
Received for publication March 29, 2007.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
Abbreviations: EAV, emergent aquatic vegetation • OEW, Orlando Easterly Wetland • Pi, inorganic P • Po, organic P • SAV, submerged aquatic vegetation • TP, total phosphorus
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
The controlling factor in the effectiveness and toxicity of alum is the pH of the system. Alum itself has a pH of approximately 2.4 (Beecroft et al., 1995; Lind, 2003) and therefore tends to decrease the pH of the system it is added to. As long as the pH of the system remains between 6 and 8, insoluble polymeric Al(OH)3 will dominate (May et al., 1979) and P inactivation results. However, if the pH decreases to between 4 and 6, soluble intermediates will occur releasing bound P, and below pH 4 and above pH 8, soluble Al3+ dominates which may result in aluminum toxicity (Cooke et al., 1993; Ma et al., 2003).
Phosphorus can be mobilized from flooded soil/sediment under continuously flooded conditions (Malecki et al., 2004) and under fluctuating water levels (Corstanje and Reddy, 2004; Bostic and White, 2007), underscoring the need for management alternatives to increase P sequestration in wetland soils. While alum has been used for P inactivation in eutrophic lakes since 1968 (Blomquist et al., 1971) there has been little research done on its potential effectiveness in aging treatment wetlands with reduced P sorption capacities (Simon, 2003; DB Environmental, Inc., 2004; Brown, 2007). There is also a lack of information on the impact of increased Al concentrations on the biomass and activity of the microbial community, and therefore overall nutrient cycling of alum-treated ecosystems.
The rate of microbial activity and structure of the microbial community is largely dependent on environmental factors. Both size and activity of the microbial pool influences the nutrient removal of a wetland (White and Reddy, 1999, 2003) as well as the removal of other contaminants (White et al., 2006a). Microbes are generally sensitive to both soil acidity (Degens et al., 2001) and soluble Al (Illmer et al., 1995; Robert, 1995; Pina and Cervantes, 1996). The microbial biomass, therefore, has the potential of being a sensitive indicator of impact to soil nutrient dynamics (Powlson and Jenkinson, 1981) due to alum application. This is due to the close relationship between microbial biomass nutrients and levels of mineralizable nutrients available in the soil (Jenkinson and Ladd, 1981; Illmer et al., 1995; Gutknecht et al., 2006).
An alum study by Connor and Martin (1989) on a shallow New Hampshire lake suggested that the dissolved oxygen (DO) concentrations within the lake may have been affected by suppression in activity or reduction in the population of microorganisms; however, neither was measured. A laboratory soil core study utilizing various Al amendments determined an inverse relationship existed between Al dose and microbial biomass and activity (Brown, 2007). These studies suggest that the size and activity of the microbial pool, with respect to alum concentration, is critical to understanding changes in P cycling within a treatment wetland in response to alum addition. Additionally, the characterization of soil P is necessary to determine shifts in exchangeable metal oxide and hydroxide, and organically bound P pools due to alum application. Alum may not only influence the P mineralization rate, but also nutrient availability to aquatic macrophytes by binding bioavailable P in the soil. The aquatic macrophytes selected for use in treatment wetlands can significantly impact the P treatment efficiency and overall nutrient budget (Tanner, 1996; Reckhow and Chapra, 1999; Thiebaut and Muller, 2000; Allen et al., 2002). Macrophyte species composition can affect the soil biogeochemistry as well (Wigand et al., 1997; White et al., 2004, 2006b; Neubauer et al., 2005).
The overall goal of this study was to investigate the effects of a slow drip alum application on the physicochemical soil characteristics, as well as the microbial biomass pool size and activity of treatment wetlands dominated by different aquatic macrophytes. Our hypotheses of this study were that alum application would (i) decrease the soil pH, (ii) increase soil Al concentrations and Al-bound P, and (iii) negatively affect soil microbial biomass pool size and activity. Additionally, we hypothesized that the impacts of alum would vary due to the type of vegetation present.
|
Materials and Methods |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
The 494-ha wetland rests on a 664-ha piece of land located 3.2 km west of the main channel of the St. Johns River (SJR). Historically, the land had been part of the riparian wetland adjacent to the SJR, but was drained for pasture by a cattle ranch around the turn of the last century (Burney et al., 1989). The site has a natural topographic gradient of 0.2% downward from west to east allowing water to flow by gravity through a series of cells with an average elevation drop across each cell of approximately 1 m (Martinez and Wise, 2003) (Fig. 1 ). Water exits the wetland through a weir control structure and flows into a receiving ditch. From there water can flow directly to the SJR or by sheet flow through Seminole Ranch, a natural marsh adjacent to the OEW owned by the St. Johns River Water Management District (SJRWMD).
|
The overall average influent total phosphorus (TP) concentration from 1988 to 2005 was 0.22 mg L–1; however, annual inflow TP concentrations range from 0.02 to 3.30 mg L–1 during the same time period. Since its inception, the OEW has exceeded performance expectations. The TP discharge permit limit established by the FDEP is 0.2 mg L–1 (Wang et al., 2006). From 1988 to 1995 the average TP discharged was 0.07 mg L–1 (Sees and Turner, 1997); however, TP values are considerably higher from December to February in recent years (Wang et al., 2006).
Mesocosm Establishment
Eighteen circular mesocosms with a surface area of 1.86 m2 and depth of 0.88 m were established in June 2004. Triplicate experimental and control units were planted with either Typha spp., Schoenoplectus californicus, or submerged aquatic vegetation (SAV) (Najas guadalupensis-dominated) utilizing a randomized block design. These three vegetation types were selected based on their dominance within the OEW. Approximately 0.3 m of homogenized soil was added to each mesocosm. The soil used to initiate the mesocosms originated from a dredged spoil pile originating from cell 1, 3, 4, 7, and 8 of the OEW (Fig. 1). Soil used had a TP concentration of 111 ± 13.4 mg kg–1, 570 ± 42.5 mg kg–1 oxalate-extractable Al, 19.9% ± 0.02 moisture content, and soil pH of 5.21 ± 0.28.
Each mesocosm contained a polyvinyl chloride (PVC) drain which permitted control of water levels to within ±3 cm. The water flowing through the mesocosms originated from the outflow of cell 15 and was pumped to a head tank and distributed (on a 1 h timer) via gravity to each mesocosm at a rate of 360 L d–1 with a hydraulic loading rate of 9.6 cm d–1, providing a retention time of approximately 9.5 d, similar to the retention time of the SAV cells within the OEW.
On 30 June 2004, the mesocosms were planted to begin the 5 mo grow-in/stabilization period. The six SAV mesocosms were established with 4.05 ± 0.06 kg WW/mesocosm, or 2.18 kg WW m–2 Najas guadalupensis harvested onsite. The stocking density for the Typha spp. and Schoenoplectus californicus mesocosms was 15 plants/mesocosm, averaging 1.88 kg WW m–2 Typha spp., and 1.04 kg WW m–2 Schoenoplectus californicus both harvested onsite as well. Stocking densities were representative of presence in OEW cells (DB Environmental, Inc., 2004). A 53 cm water column was maintained in the SAV mesocosms throughout the entire grow-in, while in the emergent mesocosms, a 28 cm water column was maintained for the first month to allow the plants to establish before raising the water column to 53 cm for the remainder of the study.
Experiment Initiation
On 1 Dec. 2004, liquid alum was pumped via peristaltic pumps on timers to the designated mesocosms through black polyethylene tubing at a rate of 0.91 g Al m–2 d–1 for 3 mo resulting in a total addition of 76.8 g Al m–2. Soil physicochemical and microbial characteristics were determined from a core collected from each mesocosm on initiation and completion of the study. Cores were sectioned into 0- to 5-cm and 5- to 10-cm intervals in the field and placed in labeled Ziploc bags in coolers of ice for transportation back to the laboratory. Samples were then transferred to polyethylene containers and stored at 4°C for analysis.
Laboratory Analysis
The following physicochemical parameters were measured on the sectioned soil samples: pH, bulk density (Blake and Hartge, 1986), mass loss on ignition (LOI), TP and total Al (AlT), oxalate-extractable Al (Alox) (McKeague and Day, 1966), 1 mol L–1 HCl-extractable metals, microbial biomass P (MBP), soil oxygen demand (SOD) (APHA, 1998), potentially mineralizable P (PMP), and inorganic P fractionation.. The amount of Al present in crystalline form was calculated as the difference between the AlT and Alox (Dolui and Chakraborty, 1998). Acid-extractable Ca, Mg, Fe, and Al concentrations were determined from oven-dried soil treated with 25 mL of 1.0 mol L–1 HCl and placed on a reciprocal shaker for 3 h. The supernatant was filtered through 0.45-µm membrane filters and analyzed by inductively coupled argon plasma spectrometry (Method 200.7, USEPA, 1993; DeBusk et al., 1994). To quantify the microbial biomass presence, MBP was determined by a 24 h chloroform fumigation-extraction (CFE) technique (Hedley and Stewart, 1982). The PMP rates were determined using an anaerobic, waterlogged incubation at 40°C (Chua, 2000; Malecki-Brown and White, unpublished data, 2007). Both SOD and PMP rates were used as measures of microbial activity.
Total P analysis involved combustion of 0.5 g oven-dried subsamples at 550°C for 4 h in a muffle furnace followed by dissolution of the ash in 6 mol L–1 HCl on a hot plate (Andersen, 1976). Total P was analyzed using an automated ascorbic acid method (Method 365.4, USEPA, 1993) while AlT was determined by inductively coupled argon plasma spectrometry (model Vista MPX CCD simultaneous ICP–OES manufactured by Varian, Inc., Walnut Creek, CA; Method 200.7, USEPA, 1993). Ash content was calculated to determine mass LOI, indicating the organic matter content in the wetland soil (Lim and Jackson, 1982).
The partitioning of P into five geochemical fractions was performed using a sequential extraction method described by Reddy et al. (1998). The fractions studied were: (i) labile P, soluble in 1 mol L–1 KCl; (ii) P bound to Al and Fe, soluble in 1 mol L–1 NaOH; (iii) alkali-extractable organic P, determined by subtracting NaOH SRP from the digested NaOH TP; (iv) P associated with Ca and Mg, soluble in 0.5 mol L–1 HCl; and (v) residual organic P, soluble after ignition and acid digestion. Phosphorus availability decreases with each sequential step providing information related to its soil mobility (Luderitz and Gerlach, 2002).
Statistical Analysis
Paired t tests were used to determine significant differences (p < 0.05) between soil properties in the 0- to 5-cm and 5- to 10-cm sectioned intervals (Microsoft, 2000). Additionally, Pearson product-moment correlation coefficients between parameters were calculated (Microsoft, 2000) to determine significant relationships. Data normality was determined using the Kolmogorov-Smirnov test (Minitab 13.32, 2000) and data was transformed to fit a normal distribution (Microsoft, 2000). One-way ANOVAs and multiple comparisons by Tukey's W were used on soil parameters to determine significant (p < 0.05) effects corresponding to plant type and treatment (Minitab version 13.32; Minitab, 2000).
|
Results and Discussion |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
|
|
The ammonium oxalate extraction was used to quantify the poorly crystalline amorphous and organically bound Al oxides, hydroxides, and oxy-hydroxides in the soil (McKeague and Day, 1966; McKeague et al., 1971; Kodama and Ross, 1991), excluding all crystalline forms (Parfitt and Childs, 1988). At experiment initiation, there was no significant difference in Alox concentrations due to plant type or between the surface (Table 1) and subsurface (Table 2), averaging 595 ± 167 mg kg–1 overall. By the end of the experiment amorphous Al concentrations were significantly higher than at time 0, and by depth there was nearly twice as much Al in the 0- to 5- (1032 ± 634 mg kg–1) than 5- to10-cm layer (593 ± 200 mg kg–1) for all mesocosms. This was expected in the mesocosms receiving alum due to the surface application of Al, but it was also true of the control mesocosms due to the natural Al associated with organic matter.
There were no differences in Alox due to plant type on study completion. Similar to the AlT, however, the alum-treated mesocosms did have significantly higher Alox concentrations in the surface layer, containing more than double the amorphous Al as the controls. Specifically, the alum-treated Typha tanks had significantly higher surface Alox concentrations than the controls of all three plant types. There were also no significant differences in Alox in the 5- to 10-cm layer similar to the AlT, indicating the Al(OH)3 floc remained at the soil surface.
There was no initial difference in crystalline Al with depth; however, by the end of the study there was generally more in the surface layer than subsurface (Table 1, Table 2). Crystalline Al concentrations within the soil did not vary with plant type or treatment in either layer throughout the experiment, averaging 1483 ± 813 mg Al kg–1 overall. Using the change in ratio of Alox/AlT over time, however, the crystallization of soil Al was detected in the surface layer of the control mesocosms as indicated by a significant decrease in the Alox/AlT ratio (Mahaney et al., 1991; Bera et al., 2005). This suggests that the Al hydroxide floc at the surface of the alum-treated mesocosms may hinder natural soil development due to an increased amount of ligand-bound amorphous material which inhibits mineralogical crystallization (Huang et al., 2002).
Similar to Alox, soil pH did not differ between surface and subsurface layers (Table 1, Table 2), nor between mesocosms at study initiation. By the end of the experiment, however, there was a significant difference in soil pH due to plant and treatment type. Corresponding to the high Alox in the surface soil of the alum-treated Typha tanks, pH was significantly lower than in the control mesocosms. Unlike the trend in Alox, however, the subsurface soil pH of the alum-treated Typha tanks was also significantly lower than the controls. This suggests that while the floc itself may have not penetrated down through the soil profile, the protons (H+) released as the Al(OH)3 floc dissolved under low pH conditions (3.3–4.5) may have equilibrated with the subsurface, reducing pH in the 5 to 10 cm soil layer as well (Huang et al., 2002). The pH was inversely correlated to the Alox in both the surface (p < 0.01) and subsurface (p < 0.05) of the Typha mesocosms while this relationship was not evident in the Schoenoplectus or SAV mesocosms.
Surface and subsurface soil pH of the alum-treated SAV mesocosms remained unimpacted by the increased Alox. The thick stands of SAV may have intercepted the settling alum floc resulting in irregular alum coverage on the soil surface. While this may have reduced impacts on the soil biogeochemistry, it could also in turn lead to reduced alum effectiveness (Welch and Schrieve, 1994; Welch and Cooke, 1999). In the Schoenoplectus tanks the surface soil pH also remained relatively constant, similar to the SAV, but subsurface pH within both the control and alum-treated mesocosms significantly increased by study end. Convergence of soil pH to neutral is a typical response associated with soils when they are flooded (Ponnamperuma, 1972; Anderson et al., 2005).
Generally all HCl-extractable metals were higher in the surface than subsurface, similar to TP and LOI, and there were no differences due to plant type throughout the study (Table 1, Table 2). Calcium was the dominant HCl-extractable metal at time 0 in both the soil surface and subsurface for all plant types making up 72% of the metals extracted on average, which was significantly higher than the Al, Fe, and Mg extracted. There was significantly more Fe (17%) and Al (10%) than Mg (1%) present in the mesocosms as well. By the completion of the study there was significantly more Ca in the surface layer of the alum-treated mesocosms while in the subsurface layer Ca was significantly lower than in the controls. This increased Ca may be associated with a temporary gypsum formation at the floc/soil interface as the excess dissociated sulfate ions from alum addition bind with available Ca in the soil (L.M. Malecki and W. Harris, unpublished data, 2003). Calcium remained the dominant acid-extractable metal composing 73% of the metals extracted compared to 15% Fe, 11% Al, and 1% Mg and can also provide significant buffering to minimize pH effects from alum additions.
Soil Microbial Characteristics
There were no significant differences in MBP with depth or between macrophyte type initially (Table 3
, Table 4
). By the end of the study there was a significant decrease in microbial biomass in all mesocosms which was attributed to the cold winter season during which the experiment was performed. However, the emergent Schoenoplectus and Typha mesocosms treated with alum had significantly lower MBP than the controls in both the surface (66% lower) and subsurface (50% lower) soil layers. These results are similar to the pH findings within the Typha mesocosms, indicating once again that while the floc itself remained at the surface, the impact of alum application penetrated down through the soil profile. Additionally, similar to soil pH, MBP in the emergent tanks was inversely related to the Alox present in both soil layers (p < 0.05) while this relationship did not exist in the SAV mesocosms.
|
|
Microbial activity, as indicated by SOD rates, was significantly greater in the surface than subsurface layer of the SAV and Typha mesocosms throughout the study, while there were no differences in PMP rates with depth due to high data variability (Table 3, Table 4). After 3 mo of alum application, however, there was a significant treatment effect. Both parameters indicated significantly greater microbial activity in the 0- to 5-cm layer of the emergent planted control mesocosms than in those treated with alum. This suppression in activity corresponds to the significant decrease in microbial biomass within the Schoenoplectus and Typha mesocosms due to alum application.
There were no significant differences in SOD microbial activity due to treatment or plant type in the subsurface layer. The alum-treated Typha mesocosms continued to have significantly lower PMP rates than controls in the subsurface; however, they were most likely attributed to the continued low soil pH discussed previously. The surface soil PMP rates were positively correlated (p < 0.05) with soil pH, similar to the MBP concentrations. This suggests once again that alum indirectly impacted both the microbial biomass and activity due to the sensitivity of microbes to the increased soil acidity (Degens et al., 2001).
Soil Phosphorus Forms
The characterization of soil P was examined to assess how alum application may change soil P availability in treatment wetlands. Identifying the dominant P fractions present can be conducive to further understanding of P cycling within the OEW during winter periods of low treatment efficiency. Organic and inorganic P were essentially equivalent in all mesocosms averaging 45% inorganic P (Pi) (Table 5
) and 55% organic P (Po) (Table 6
) in the 0- to 5-cm layer at the start of the study. By the completion of the study there was a significant increase in the total P pool, with organic P (67%) dominating the inorganic P (33%) in the surface soil; however, there were no significant differences with treatment or plant type. Generally all P fractions were greater in the surface layer than in the subsurface (Table 7
, Table 8
), similar to the TP values previously discussed. The overall decrease in P storage capacity with depth may be attributed to the decrease in metal concentrations (Al, Ca, Fe, Mg) with depth (Table 1, Table 2).
|
|
|
|
|
The initial subsurface 5 to 10 cm soil had nearly equal amounts of organic and inorganic P similar to the surface layer averaging 52% Po (Table 7) and 48% Pi (Table 8), while it was dominated by organic P by the end of the study averaging 64% Po versus 36% Pi. Labile P once again represented 0.1 to 0.4% of the total P, identical to the surface soil (Fig. 2). The Al- and Fe-bound P made up 11 to 28% of the subsurface P pool initially, and only increased slightly in the alum-treated mesocosms (15 to 29%). This indicates once again that the alum floc remained primarily at the surface of the soil where substantial shifts in the NaOH Pi and Po pools were evident. The dominant inorganic P form in the subsurface layer of most mesocosms was the Ca- and Mg-bound P which ranged from 9 to 50% of the total P pool which, similar to the surface soil, can be primarily attributed to the significantly high Ca concentration (Table 2). The organic acid and bacterial P fraction made up 15 to 32% of the total P throughout the study in the subsurface, while the recalcitrant Po composed 22 to 36% of the total P.
|
Conclusions |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
The increase in amorphous Alox within the soil surface layer of alum-treated Typha mesocosms was directly associated with a decrease in soil pH throughout the upper 10 cm. The increased Alox concentration within the Schoenoplectus as well as reduced pH within Typha mesocosms, were correlated to decreased microbial biomass as well as activity. The use of alum may, therefore, impair P mineralization and overall cycling within emergent systems in the short term. In contrast, the pH and microbial community within soil of the SAV mesocosms was unimpacted; however, further research is needed to determine if the dense vegetation intercepted the settling alum floc, possibly impacting SAV growth or nutrient uptake rather than affecting the soil. Depending on soil properties, some systems may require a buffer agent to minimize the drop in pH due to alum additions, which adds cost and complexity to potential management implementation. The pH drop appears to be short lived in the water column of many lakes as well as in our wetland study (Malecki-Brown and White, unpublished data, 2007). However, few studies have ever documented the effect of the alum addition to the sediment/soil pH. Additionally, the specific mechanism on the negative impacts of low-dose alum application on the microbial community needs to be determined, whether from pH, associated aluminum toxicity, or decreased availability of bioavailable P in the soil. If the decrease in microbial activity is related to pH or related Al toxicity, then a pH buffer should neutralize this effect. It is critical that treatment wetland managers assess the short and possibly long-term implications when using alum in constructed wetland systems to maximize system performance.
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMaterials and MethodsResults and DiscussionConclusionsREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. M. Malecki-Brown and J. R. White Effect of Aluminum-Containing Amendments on Phosphorus Sequestration of Wastewater Treatment Wetland Soil Soil Sci. Soc. Am. J., May 1, 2009; 73(3): 852 - 861. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Belmont, J. R. White, and K. R. Reddy Phosphorus Sorption and Potential Phosphorus Storage in Sediments of Lake Istokpoga and the Upper Chain of Lakes, Florida, USA J. Environ. Qual., March 25, 2009; 38(3): 987 - 996. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M. Malecki-Brown, J. R. White, and M. Sees Alum Application to Improve Water Quality in a Municipal Wastewater Treatment Wetland J. Environ. Qual., February 25, 2009; 38(2): 814 - 821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. R. White, L. M. Gardner, M. Sees, and R. Corstanje The Short-Term Effects of Prescribed Burning on Biomass Removal and the Release of Nitrogen and Phosphorus in a Treatment Wetland J. Environ. Qual., October 23, 2008; 37(6): 2386 - 2391. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 | |||
