Introduction
Phosphorus (P) enrichment of freshwater systems can stimulate increased growth of aquatic plants and algae, leading to accelerated eutrophication, when P is the main growth-limiting nutrient (Douglas et al. 2004). Improvements in the treatment of point-source pollution have led to an increase in the relative importance of non-point source pollution on declining surface water quality in New Zealand (Scarsbrook 2006). Consequently, resource managers from regulatory authorities (Regional councils) are placing greater emphasis on the control of non-point source pollution associated with intensive agriculture. Increased stocking rates on pastoral farms contribute to increased transfer of nitrogen (N) and P from grazed pastures to surface water bodies (Monaghan et al. 2006). Accelerated P loss in surface runoff and in mole and pipe drainage from dairy farms has been reported and quantified in several research studies (Sharpley and Syers 1979; Nash and Murdoch 1997; Monaghan et al. 2002; Houlbrooke et al. 2003). These P losses can be particularly large from land on which farm dairy effluent (FDE) is applied (Houlbrooke et al. 2004b). When poorly treated FDE is applied to artificially drained soils there is the risk that FDE will move rapidly through preferential flow paths in the soil to mole channels and then to drainage pipes discharging to surface waters (Houlbrooke et al. 2004a). Phosphorus losses in rainfall-induced winter drainage have also been shown to be higher from dairy farm areas receiving FDE than non-effluent areas (Houlbrooke et al. 2003).
Mole and pipe drainage systems are commonly employed by intensive agriculture in temperate regions of the world to overcome the imperfect drainage caused by fine-textured subsoils. In New Zealand, mole and pipe drainage systems are widely used, including for intensive dairying in the Manawatu, Northland, Southland, and Otago regions. In situations where drainage waters are excessively enriched in P, it has been proposed that the use of high P-adsorbing materials, as fill in drainage ditches or in underground mole and pipe drainage networks, could be effective in trapping P before it can enter surface waters (Heal et al. 2004; Monaghan et al. 2005).
Much of the research conducted on the capacity of P-adsorbing materials to remove P from water has focused on their use in constructed wetland systems (Mann 1997; Sakadevan and Bavor 1998; Tanner et al. 1999; Brix et al. 2001; Gruneberg and Kern 2001; Pant et al. 2001; Naylor et al. 2003; Heal et al. 2005; Drizo et al. 2006). Constructed wetlands, with subsurface flow, are widely used throughout the world to treat a wide variety of wastewaters (Del Bubba et al. 2003). However, constructed wetlands generally have a greater potential to remove N, by biological denitrification, than they do P (Arias et al. 2001). The main methods of P removal from wastewaters in constructed wetlands are plant uptake, assimilation by microorganisms, and physico-chemical processes. Among the physico-chemical processes, P adsorption by soil/media and precipitation reactions play an important role (Del Bubba et al. 2003). Phosphorus removal efficiency is often high initially and then decreases after some time as the P adsorption capacity of the media is exhausted. Accordingly, it is important to select a media with a high P adsorption capacity for sustained P removal over the long term (Arias et al. 2001).
The material properties important for influencing P adsorption include the presence of minerals with reactive iron (Fe) or aluminium (Al) hydroxide or oxide groups on their surfaces, and calcareous materials, which can promote calcium phosphate precipitation (Zhu et al. 1997; Drizo et al. 1999). Also, adsorption is controlled by the material's pH-dependent surface charge and adsorptive surface area. Materials with small particle size have large surface areas for a given volume and, therefore, the potential to enhance P adsorption capacity (Zhu et al. 1997). However, materials with smaller particles size, when packed in beds, also have lower hydraulic conductivity, which can lead to the occurrence of the either restricted flow or by-pass flow. The latter can result in insufficient contact between the wastewater and the media within the constructed wetland (Drizo et al. 1999). Anion competition, such as the competitive adsorption of phosphate (P[O.sub.4.sup.3-]) and sulfate (S[O.sub.4.sup.2-]), is another factor that needs consideration when evaluating the effectiveness of media to adsorb P from drainage or wastewaters (Parfitt 1982; Ryden et al. 1987).
Recently, research has focused on using high P adsorbing materials as 'active filters', which are separate filtration units attached to constructed wetlands or wastewater treatment systems (Arias et al. 2003; Shilton et al. 2005, 2006). Also, the use of P-adsorbing materials as soil amendments to reduce soil P losses, in surface runoff and drainage, has also been studied (McDowell 2004, 2005). By-product materials, such as fly ash from coal-fired power plants and steel industry slags, have particular potential for widespread application as they are produced in large quantities and often present storage or disposal problems (Douglas et al. 2004). While these materials have the advantage of being relatively inexpensive, a major drawback is their toxicity. For example, steel slags can contain concentrations of arsenic, cadmium, and mercury above the levels permitted for disposal on land (McDowell 2004). Also, McDowell (2004) demonstrated that fly ash, when applied to soil, has the potential to cause boron toxicity to plants. In some instances, by-product materials have increased P losses from soil despite containing P-adsorbing iron and aluminium compounds. For example, McDowell (2005) showed that fly ash, when added to some soils, increased soil P losses by raising soil pH from <6 to 6-7, where P is most soluble. This increase in pH was attributed to inputs of calcium and magnesium compounds contained in the fly ash (McDowell 2005).
New Zealand has abundant naturally occurring materials with high P adsorption capacities, such as soils and moderately weathered materials derived from volcanic tephra. Moderately weathered materials that contain the mineral allophane show particular promise for removing P from wastewater. Ryden and Syers (1975) demonstrated that of the tephra materials they studied, moderately weathered andesitic tephra had the greatest potential for removing P from effluent. They found that andesitic tephra from Taranaki, New Zealand, was capable of removing 80-97% of P from a synthetic P solution containing 5 mg P/g tephra.
The aim of this study was to evaluate the effectiveness of one particular type of tephra (Papakai tephra) as a P-adsorbing material for removing P from dairy farm drainage water when used in mole and pipe drainage systems.
Materials and methods
Laboratory study
Materials
The tephra used in this study, Papakai tephra, was collected from the Mangatoetoenui Quarry, which is located beside the Mangatoetoenui Stream crossing of State Highway 1 (NZMS 260, T20, 460 152) on the Tongariro Volcanic Centre (TgVC). The Papakai tephra (ash and lapilli) was taken from a depth of c. 3.45-3.75 m and represents the lowest strategraphic layer of the Papakai Formation, which is dated c. 9790-2500 years BP (Donoghue et al. 1995).
The Papakai tephra has an average pH 6.3, extractable sulfate 19.5mg S/kg (Blakemore et al. 1987), Olsen extractable P 0.9 mg …
