Artifacts Caused by Collection of Soil Solution with Passive Capillary Samplers

1979; Barbee and Brown, 1986; Grossmann and Udluft, 1991; Brandi-Dohrn et al., 1996a). The area or zone Soil solution was collected with passive capillary samplers of sampling is unknown and may change with varying (PCAPS) and zero-tension samplers (ZTS) from A horizons of forested soils. The volume and chemistry of collected solutions were moisture conditions or suction applied. In addition, sucmeasured weekly during discrete seasonal collection periods. Acidtion must be applied manually to the sampler several washed PCAPS increased alkalinity (3-fold), pH (1–3 units), and hours or days prior to solution collection. This is probconcentrations of Ca (2-fold), Na (8-fold), and Si (4-fold), relative to lematic because the magnitude of tension exerted on ZTS solutions. Aluminum concentrations were dramatically reduced soil water gradually decreases over time. Lastly, because in PCAPS compared with ZTS samples. Differences in solution chemof the small cross-sectional area of the cup, multiple istry were attributed to leaching and weathering of fiberglass wicking samplers are required to represent adequately soil varimaterial utilized in the PCAPS. In addition, PCAPS collected greater ability (Barbee and Brown, 1986). volumes (normalized by sampler area) of solution relative to ZTS in Brown et al. (1986) first introduced passive capillary weak-structured sandy loam soil due to a preponderance of matrix flow. The results indicate that the PCAPS used in this study are not samplers (PCAPS) to provide an alternative means of suitable for aqueous geochemical studies of dilute soil solutions. sampling soil water in the field. As a result of their high conductivity and the tension exerted, PCAPS collect matrix and macropore flow under both saturated and T analysis of soil solution is an essential aspect unsaturated conditions, thus providing a more represenof soil science. Soil solution is the primary mediator tative soil solution sample and improved collection effiof pedogenic processes and it influences groundwater ciency (Boll et al., 1991). The specific design of PCAPS and surface water quality. However, prior to analysis, varies slightly in the literature (Holder et al., 1991; Boll soil solution must first be isolated from the soil. Numeret al., 1991; Biddle et al., 1995; Brandi-Dohrn et al., ous in situ devices have been employed over the years 1996b) but the basic structure includes woven fiberglass to collect soil solution, including zero-tension samplers ropes (“wicks”) that create a hanging water column or ZTS (Parizek and Lane, 1970; Haines et al., 1982; to produce a tension where the wicks contact the soil. Jemison and Fox, 1992; Marques et al., 1996) and suction Collected water is then transported along the length of cup solution samplers (Parizek and Lane, 1970; Shaffer the wicks into a collection container. The cross-sectional et al., 1979; Barbee and Brown, 1986; Chorover et al., area of the samplers is easily manipulated for installa1994; Brandi-Dohrn et al., 1996a; Marques et al., 1996). tion in stony soils. Although these are the most commonly used samplers, The effects of wick material on the chemical composiboth have inherent flaws in collecting solution. tion of collected solutions is not well known. Holder et The main limitation of ZTS is that they only sample al. (1991) determined that fiberglass wicks did not adthat fraction of soil water flux occurring under saturated sorb Br, NO 3 , Cd21, ethylbenzene, toulene, naphthasoil conditions or during macropore flow. In response lene, or trichloroethylene. Boll et al. (1992) reported to hydraulic gradients, flow lines are potentially diverted minimal retardation of FD&C no.1 blue dye in wicks. around the solution sampler when soil water content is However, Biddle et al. (1995) found there was an initial below saturation (Boll et al., 1991; Jemison and Fox, contamination of solution from CO22 3 at a level of almost 1992). Another constraint of ZTS is the large cross25 mol/m3 even after washing with 5% (v/v) phosphatesectional area requirement necessary for collection effifree enzyme-surfactant, rinsing with 5% (v/v) HCl, and ciency (Radulovich and Sollins, 1987). Also, stony soils then rinsing with deionized water. After leaching the do not allow for the practical installation of large ZTS wicks with 1.5 L of distilled water, the CO22 3 concentrasuch as those used by Jemison and Fox (0.465-m2 crosstion decreased to ,80 mmol m23 and soil solutions colsectional area; 1992). lected subsequently were considered unaffected by the Although suction cup samplers may be used to collect wick material. Biddle et al. (1995) employed PCAPS to solution under unsaturated conditions, they do not proassess major ion and aqueous colloid chemistry in a field vide a reliable estimate of soil water flux (Shaffer et al., soil but they were unable to evaluate potential artifacts because no other kind of sampler was used for compariDep. of Agronomy, The Pennsylvania State Univ., 116 A.S.I. Bldg., son. Other researchers have shown the utility of PCAPS University Park, PA 16802. Received 14 June 1999. *Corresponding for measuring transport of non-reactive species (e.g., author ([email protected]). bromide and nitrate) in soils (Holder et al., 1991; Brandi-Dohrn et al., 1996b). However, their application Published in Soil Sci. Soc. Am. J. 64:1330–1336 (2000).