Ground Water Nitrate Removal in Subsoil of Forested and Mowed Riparian Buffer Zones

We studied two similar riparian sites in southern New England and examined ground water nitrate (NO~--N) removal in the subsurface of mowed (i.e., herbaceous) v . forested (i.e., woody) vegetation. Each site consisted of poorly drained, fine to medium sands and contained adjacent areas of mowed and forested vegetation. We dosed mesocosms with bromide and lSN labeled NO~-N amended ground water to simulate the shallow ground water NO~--N dynamics of riparian buffer zones. Mesocosms were composed of undisturbed, horizontal soil cores (40 cm long, 15 cm diam.) extracted from seasonally saturated subsoil We observed substantial ground water NO~-N removal and denitrification at all locations. Ground water NO~-N removal rates were significantly correlated with carbon-enriched patches of organic matter. This correlation supports previous work that patches function as hotspots of microbial ctivity in the subsoil. Within each site, we found no significant difference in ground water NO~--N removal rates in the subsoil of forested and mowed areas and we noted tree roots throughout the subsoil of the mowed areas. We found that ground water NO~-N removal rates differed significantly between similar sites. We caution against ascribing specific ground water NO~--N removal rates to different riparian aboveground vegetation types without recognizing the importance of site differences, e.g., water table dynamics, land use legacy and adjacent vegetation. Riparian zones composed of a mix of forested and mowed vegetation, common in agroforestry and suburban land uses, may remove substantial amounts of ground water NO~--N. N ITRATE (NO£) has been linked to the eutrophication of coastal waters (Ryther and Dunstan, 1971; Howarth et al., 1996; Jordan et al., 1997). Our ability to manage the export of NO~--N from coastal watersheds is limited by a lack of understanding of the various processes that retain or remove NO~--N in terrestrial landscapes (Howarth et al., 1996; Jordan et al., 1997). Scientific consensus exists that riparian zones, transition areas between uplands and surface waters (Gregory et al., 1991), can be significant sinks for ground water NO~--N (Lowrance et al., 1984; Peterjohn and Correll, 1984; Hill, 1996; Gilliam et al., 1997; Jordan et al., 1997; Lowrance, 1997). There is considerable uncertainty, however, regarding the site characteristics that promote substantial ground water NO~--N removal in riparian zones and the influence of different types of riparian vegetation cover on ground water NO~--N removal (Korom, 1992; Gilliam, 1994; Hill, 1996; Gilliam et al., 1997; Correll, 1997). Substantial spatial variability in ground water NO~--N removal often occurs within riparian zones. In our previous work (Groffman et al., 1992; Simmons et K.L. Addy and A.J. Gold, Dep. of Natural Resources Science, Univ. of Rhode Island, 210B Woodward Hall, Kingston, RI 02881; and P.M. Groffman and P.A. Jacinthe, Inst. of Ecosystem Studies, Box AB, Millbrook, NY 12545. *Corresponding author ([email protected]). PubLished in J. Environ. Qual. 28:962-970 (1999). al., 1992; Hanson et al., 1994; Nelson et al., 1995; Gold et al., 1998; Jacinthe et al., 1998), we found that poorly drained (PD) forested soils have a high capacity for ground water NO;--N removal, much higher than adjacent moderately well drained (MWD) forested soils. observed that different drainage classes in sandy soils have distinct water table dynamics; dormant season PD water tables generally rise within 0.3 m of the surface, whereas MWD water tables only come within 0.8 m of the surface. In addition, Gold et al. (1998) and Jacinthe et al. (1998) found that PD shallow aquifer material contained patches of organic matter in the C horizon, whereas MWD soils contained no patches in the C horizon. These patches function as hotspots of microbial activity (Gold et al., 1998; Jacinthe et al., 1998). Soils with a greater proportion of patch material in the saturated zone may serve as important sinks for ground water NO~-N within the landscape. Most recent riparian zone research has focused on forested areas (Lowrance et al., 1984; Peterjohn and Correll, 1984; Ambus and Lowrance, 1991; Lowrance, 1992; Simmons et al., 1992; Hanson et al., 1994; Nelson et al., 1995; Groffman et al., 1996; Starr et al., 1996; Gold et al., 1998; Jacinthe et al., 1998). Fewer studies have examined differences in NO~--N removal between forested and nonforested riparian zones. Some studies (Osborne and Kovacic, 1993; Hubbard and Lowrance, 1997) found grassed riparian sites had lower NO;--N removal rates than forested riparian sites, but other studies (Haycock and Burt, 1993; Haycock and Pinay, 1993; Lowrance t al., 1995; Schnabel et al., 1996; Correll et al., 1997) observed substantial removal in grassed riparian sites. More studies are needed to determine the relative importance of site characteristics vs. vegetation cover on ground water NO~-N removal in riparian subsoils. The goal of our study was to examine the effects of vegetation type on ground water NO;--N removal rates in the seasonally saturated subsoil of PD riparian zones. We used saturated mesocosms for our analyses. These mesocosms were undisturbed soil cores of about 11 kg of soil that were extracted horizontally from the subsoil. These cores were obtained from adjacent forested (i.e., woody) and mowed (i.e., herbaceous) areas at two different riparian sites. The mesocosms were continuously dosed with bromide (Br-) and ~SN labeled NO~--N amended ground water. Because the ground water NOA--N in these saturated mesocosms was not subjected to plant uptake, only microbially mediated NO;--N removal processes were examined. The specific objectives of this study were: (i) to compare ground water NO£-N Abbreviations: PD, poorly drained; MWD, moderately well drained; DEA, denitrification enzyme activity; DO, dissolved oxygen.