Science of the Total Environment 494–495 (2014) 329–336

Contents lists available at ScienceDirect

Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv

Influence of urban land development and subsequent soil rehabilitation on soil aggregates, carbon, and hydraulic conductivity Yujuan Chen a, Susan D. Day a,b,⁎, Abbey F. Wick c, Kevin J. McGuire a a b c

Department of Forest Resources & Environmental Conservation, Virginia Tech, Blacksburg, VA 24061, USA Department of Horticulture, Virginia Tech, Blacksburg, VA 24061, USA Department of Soil Science, North Dakota State University, Fargo, ND 58108, USA

H I G H L I G H T S

G R A P H I C A L

A B S T R A C T

• Urban land development reduces soil macroaggregates and permeability. • Can subsurface soil rehabilitation with compost mitigate these effects? • Soil rehabilitation does not measurably enhance aggregate formation within 5 years. • Soil rehabilitation does improve subsurface hydraulic conductivity. • Urban soil ecosystem service provision is strongly management dependent.

a r t i c l e

i n f o

Article history: Received 19 March 2014 Received in revised form 20 June 2014 Accepted 21 June 2014 Available online xxxx Editor: Eddy Y. Zeng Keywords: Aggregate-associated carbon Compost Land use change Runoff Soil compaction Soil restoration

a b s t r a c t Urban land use change is associated with decreased soil-mediated ecosystem services, including stormwater runoff mitigation and carbon (C) sequestration. To better understand soil structure formation over time and the effects of land use change on surface and subsurface hydrology, we quantified the effects of urban land development and subsequent soil rehabilitation on soil aggregate size distribution and aggregate-associated C and their links to soil hydraulic conductivity. Four treatments [typical practice (A horizon removed, subsoil compacted, A horizon partially replaced), enhanced topsoil (same as typical practice plus tillage), post-development rehabilitated soils (compost incorporation to 60-cm depth in subsoil; A horizon partially replaced plus tillage), and pre-development (undisturbed) soils] were applied to 24 plots in Virginia, USA. All plots were planted with five tree species. After five years, undisturbed surface soils had 26 to 48% higher levels of macroaggregation and 12 to 62% greater macroaggregate-associated C pools than those disturbed by urban land development regardless of whether they were stockpiled and replaced, or tilled. Little difference in aggregate size distribution was observed among treatments in subsurface soils, although rehabilitated soils had the greatest macroaggregate-associated C concentrations and pool sizes. Rehabilitated soils had 48 to 171% greater macroaggregate-associated C pool than the other three treatments. Surface hydraulic conductivity was not affected by soil treatment (ranging from 0.4 to 2.3 cm h−1). In deeper regions, post-development rehabilitated soils had about twice the saturated hydraulic conductivity (14.8 and 6.3 cm h−1 at 10–25 cm and 25–40 cm, respectively) of undisturbed soils and approximately 6–11 times that of soils subjected to typical land development practices. Despite limited effects on soil aggregation, rehabilitation

⁎ Corresponding author at: Forest Resources & Environmental Conservation (MC0324), Cheatham Hall, Rm 310, 310 West Campus Dr, Blacksburg, VA 24061, USA. Tel.: +1 540 231 7264; fax: +1 540 231 3698. E-mail address: [email protected] (S.D. Day).

http://dx.doi.org/10.1016/j.scitotenv.2014.06.099 0048-9697/© 2014 Elsevier B.V. All rights reserved.

330

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

that includes deep compost incorporation and breaking of compacted subsurface layers has strong potential as a tool for urban stormwater mitigation and soil management should be explicitly considered in urban stormwater policy. © 2014 Elsevier B.V. All rights reserved.

1. Introduction By 2030, urban land cover will increase by 1.2 million km2, nearly tripling the global urban land area extant in 2000 (Seto et al., 2012). As part of the initial disturbance resulting from conversion of rural land to urban land uses, soils are typically degraded by a wide range of modifications including vegetation clearing, topsoil removal, grading, and compaction. These practices adversely influence soil physical characteristics desirable for ecosystem service provision, and consequently urban soils may have increased bulk density (Jim, 1993), disrupted aggregation (Jim, 1998b), and reduced porosity (Alaoui et al., 2011). Ultimately urban soil degradation leads to the loss of critical soilmediated ecosystem services such as net primary productivity (Milesi et al., 2003), carbon (C) storage (Chen et al., 2013), and stormwater mitigation (Pitt et al., 2008). These and other soil ecosystem services, in particular water-related services, are closely linked to soil structure. For example, soil compaction resulting from urbanization can alter soil aggregate arrangement, pore space, and consequently change soil hydraulic properties (Alaoui et al., 2011). The high proportion of impervious surfaces found in urbanized areas, including the nearly impervious surfaces resulting from soil degradation (Gregory et al., 2006), can lead to flooding downstream, rapidly fluctuating stream levels, and degraded surface water quality (Leopold, 1968; Paul and Meyer, 2001). Urban soil hydraulic properties have been studied from landscape to aggregate scales. At the landscape scale, several studies report infiltration rate reduction in compacted urban soils (Gregory et al., 2006; Woltemade, 2010). In a simulation study, Berthier et al. (2004) demonstrated that soil (i.e., not paved or covered with other impervious surfaces) contributed an average of 14% of the total runoff volume at the small catchment scale although the per-event percentage varied by storm intensity. There is increasing interest in environmentally sensitive stormwater management practices that take into account the influence of site and soil variables on water movement (Pitt and Clark, 2008). A wide range of best management practices have been developed (e.g., Bartens et al., 2008; Collins et al., 2010; Xiao and McPherson, 2011) including compost amendment application (Olson et al., 2013; Pitt et al., 1999) that aim to alleviate soil compaction and facilitate water movement and storage through the soil profile. Soil compaction also can influence hydraulic properties in the aggregates themselves, depending on interactions between compaction level, aggregate size and depth (Lipiec et al., 2009). Moreover, the contacts between aggregates control unsaturated water flow (Carminati and Flühler, 2009). Aggregate size distribution can also influence water movement in soils. For example, Abu-Sharar et al. (1987) observed saturated hydraulic conductivity (Ksat) reduction resulting from aggregate break down and related macropore loss. In addition to their influence on water flow, aggregates also physically protect soil organic matter (Tisdall and Oades, 1982) and indirectly affect soil C dynamics by regulating microbial activity, water, oxygen, and nutrients in soils (Six et al., 2004). Soil aggregates are sensitive to management practices (Six et al., 1998), but do have the potential to recover after disturbance (Kay, 1998; Wick et al., 2009a). Management that results in aggregate breakdown may ultimately lead to soil C loss and increased stormwater runoff. Although, Jim (1998a) found that the proportion of water stable aggregates in highly disturbed roadside soils in Hong Kong was very low; to our knowledge no other studies have specifically explored soil aggregate size distribution in urban areas or the response of aggregates to urban disturbance and management practices and subsequent effects on aggregate-mediated ecosystem services.

Because organic material is a significant component of the binding agents that form aggregates (Six et al., 2004), it has been postulated that enhancements to soil aggregation drive increases in soil permeability resulting from compost amendment. Previous studies show that soil organic amendments can improve water holding capacity (Khaleel et al., 1981), increase water retention, especially in sandy soil (Rawls et al., 2003), and produce higher infiltration rates (Boyle et al., 1989; Brown and Cotton, 2011; Martens and Frankenberger, 1992). In urban systems, there is also considerable interest in rehabilitating urban soils with organic amendments to restore some of the ecosystem services diminished during urban land development (Cogger, 2005; Sloan et al., 2012) and compost amendments have been demonstrated to increase C storage (Chen et al., 2013), increase infiltration (Pitt et al., 1999), and improve net primary productivity (De Lucia et al., 2013; Layman, 2010). However, use of a soil amendment is typically accompanied by physical manipulation of the soil to facilitate incorporation and increases in soil permeability may be linked to factors other than increased aggregation. The majority of these studies in urban systems only address surface applications or shallow incorporation of organic amendments (e.g., Cogger, 2005 and Sloan et al., 2012). However, deep tillage accompanied by compost amendment has potential to loosen subsurface soils that are typically compacted during urban development and land use change and thus improve infiltration rates. In addition to its relation to soil structure, organic matter incorporation could also indirectly affect site hydrologic processes through increases in above- and below-ground plant growth. Urban tree canopy cover, for example, can help reduce peak discharge and stormwater runoff (Sanders, 1986) through rainfall interception (Xiao et al., 2000) as well as by increasing the permeability of the soil through root channels (Bartens et al., 2008; Johnson and Lehmann, 2006). Because of the critical role of soil aggregates in soil hydraulic properties, as well as in protecting soil C and improving soil productivity, restoring soil structure by enhancing aggregation is highly desirable and is the focus of many urban soil management practices that employ organic amendments to rehabilitate degraded soils. Whether and how quickly such rehabilitation can alter soil aggregation processes and effect long-term changes in hydraulic conductivity in soils disturbed by urban development, however, is not known. In our study, we used controlled experimental plots to address the effects of urban land development and subsequent soil rehabilitation on soil structure and permeability. We investigated whether rehabilitating degraded urban soils via deep tillage and compost incorporation plus tree planting can increase soil hydraulic conductivity and if effects are related to changes in soil aggregation or other factors. Soil profile rebuilding (PR) is a technique to rehabilitate degraded urban soils post-development to better support vegetation. In an earlier study (Chen et al., 2013), we found that PR resulted in greater C sequestration including increases to the aggregate-protected C pool, especially in subsurface soils. This suggests that rehabilitation affected the aggregate-organic matter complex, but it is unclear if this is because of improved aggregation, increased aggregate-associated C concentration, or a combination. Thus our objectives were to: (1) quantify the effects of urban land use development on soil aggregate size distribution, aggregate-associated carbon, and hydraulic conductivity (2) explore whether post-development soil rehabilitation mitigates these effects (3) determine the relationship between changes in soil structure and hydraulic conductivity.

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

2. Materials and methods 2.1. Site information The study site in Montgomery County, Virginia (N 37.200267, W 80.586493) contains two closely related loamy soils, Shottower loam (fine, kaolinitic, mesic Typic Paleudults) and Slabtown loam (fine– loamy, mixed, mesic Aquic Paleudalfs) and was in agricultural use (pasture) prior to installation of the experiment. Four soil treatments were installed in a completely randomized experimental design between May and November in 2007 (6 replications × 4 soil treatments) in plots measuring 4.6 × 18.3 m. With the exception of undisturbed control plots, all plots were subjected to pre-treatment replicating the scraping and compaction typical of land development: the A horizon was removed and stockpiled nearby, and the exposed subsoil surface was compacted with a 4808 kg ride-on sheep's foot vibrating compactor (Model SD45D, Ingersoll Rand) in 8 passes to an average subgrade bulk density at 5–10 cm of 1.98 Mg m−3. In addition to the undisturbed control (UN), treatments included: typical practice (TP), stockpiled topsoil applied to the compacted subsoil to a depth of 10 cm; enhanced topsoil (ET), same as TP, except after soil application, the site was rototilled to a 12–15 cm depth; and profile rebuilding (PR) as the soil rehabilitation treatment. Application of stockpiled topsoil was accomplished by placing topsoil on site with a front-end loader followed by dragging a specially fabricated box across the soil surface to spread soil to exactly 10-cm depth. Topsoil was placed in one layer and not compacted. The PR treatment comprises four steps: 10 cm of leaf compost (C/N ratio 15.0) applied to the surface; deep tillage with a backhoe to a 60-cm depth that breaks soil into clods of 30-cm diameter or less; replacement of 10 cm topsoil and tilling. Existing vegetation on all plots was killed with herbicide (glyphosate). The control (UN), representing agricultural land use, allows the comparison between pre- and post-urban land development soils as well as serving as a benchmark for the effects of soil rehabilitation. The TP treatment is similar to typical urban land development practices currently employed in the United States. The PR treatment is a soil rehabilitation practice intended to loosen compacted soil and accelerate aggregate formation over time, while ET is a modification of TP sometimes specified by site designers to disrupt the interface between compacted subgrades and reapplied topsoil. In spring 2008, five landscape-sized trees (approximately 1.5–2.0 m height), one of each of the species listed below, were planted 3.7 m apart in all study plots regardless of treatment. Trees were installed in each plot in a single row with in-row position randomly assigned. Species were: Acer rubrum L., Quercus bicolor Willd., Quercus macrocarpa Michx., Ulmus ‘Morton’ (Accolade®) (Ulmus. japonica (Rehd.) Sarg. × U. wilsoniana Schneid.), and Prunus ‘First Lady’ (Prunus. ×incam Ingram ex R. Olsen & Whittemore ‘Okamé’ × Prunus. campanulata Maximowicz). All plots were covered with polypropylene net and straw erosion control blankets for the first two years until soil was stabilized. Otherwise, the soil surface was kept bare (no mulch or grass) in order to facilitate study of soil C dynamics. Weeds were controlled in all plots during the entire course of the experiment with periodic applications of glyphosate and oxyfluoren + pendimethalin. 2.2. Aggregate size distribution, aggregate-associated carbon, and soil bulk density In June of 2011 and 2012, soil samples were collected at locations equidistant from the trunk of A. rubrum trees in each plot. Four soil samples were extracted from the 0–5 cm and 5–10 cm depths at a distance of 1 m from the trunk. Two of these four sampling locations were then randomly selected and additional samples extracted at 15–30 cm depth. Soil samples were composited by depth and the amount of soil required for aggregate analysis was separated from the pooled sample and dried at room temperature. We used the wet sieving protocol described by Six et al. (1998) to determine the water stable aggregate size distribution. A

331

50 ± 0.02 g sample of the air-dried soil was submerged in deionized water for 5 min at room temperature on a 2000 μm sieve. Water stable soil aggregates (N2000 μm) were separated from the whole soils by moving the sieve up and down 50 times in 2 min. The remainder of the material (soil and water) was then passed through first a 250 μm sieve and then a 53 μm sieve to separate water stable soil macroaggregates (250–2000 μm) and microaggregates (53–250 μm) using the same procedure. The material collected from each sieve (2000, 250 and 53 μm) was dried at 55 °C until a constant weight was achieved. The fine fraction (b 53 μm) was determined by subtracting the three aggregate weights from the whole soil weight (50 ± 0.2 g). Soil mean weight diameter was calculated according to Eq. (1):     MWD ¼ MN2000μm  5 þ M 250‐2000μm  1:125 þ M 53‐250μm    0:151 þ M 53μm  0:0265

ð1Þ

where M is the proportion of the soil weight in the aggregate class with a size given in the subscript. Because sand may be separated out in the sieving procedure and sand content varies with different aggregate size classes (Elliott et al., 1991), samples were corrected for sand content according to Denef et al. (2001) using 5 g of each aggregate sample. We dispersed these samples with 0.5% sodium hexametaphosphate on a shaker for 18 h. Dispersed samples were then sieved with 250- and 53μm nested sieves for macroaggregates and microaggregates, respectively. Collected sand was dried and weighed and sand-corrected aggregate weights were determined according to Eq. (2). Sand corrected weight ¼ aggregate weight−½ðsand weight=5 gÞ  aggregate weight:

ð2Þ

Macroaggregate and microaggregate samples were analyzed for total C using dry combustion on an Elementar Variomacro CN Analyzer (Elementar Analysensysteme GmbH, Hanau, Germany). Carbon data are presented as both concentrations (g C kg−1 sand free aggregate) and pool sizes (Mg ha− 1). In order to compare C concentration across sites, we determined sand-free C by the following formulas for sand correction [Eq. (3)] for each size class (Denef et al., 2001; Wick et al., 2009a): Sand free C ¼ C  ½aggregate=1−percentage of sand

ð3Þ

To characterize soil bulk density, soil cores (diameter = 5 cm, height = 5 cm) were collected at 4 depths (2.5–7.6 cm, 15.2–20.3 cm, 30.5–35.6 cm and 50.8–55.9 cm) in the middle of each of plot using a slide hammer in July 2012. Bulk density was calculated for each sample after oven drying at 105 °C. 2.3. Hydraulic conductivity measurement Infiltration as near-saturated hydraulic conductivity (Knear) of the soil matrix was measured using a mini disk tension infiltrometer (Decagon Devices, Inc., Pullman, WA) with h =− 2 cm tension in each plot from June to July in 2012. Measurement locations were randomly selected within the central portion of each plot to avoid edge effects. Three minidisk measurements were conducted at the selected spot simultaneously. Then, a 6-cm diameter cylindrical hole was bored to 25 cm at the same measurement locations as the mini disk infiltrometer measurements to measure Ksat using a compact constant head permeameter (Ksat, Inc., Raleigh, NC) within the soil profile. A constant 15-cm head of water was maintained at the bottom of the hole and Ksat was measured from 10 to 25 cm. Similarly, a 6-cm diameter cylindrical hole was bored to 40 cm at the same location to measure Ksat from 25 to 40 cm with a 15-cm constant head of water during July and August 2012. One measurement was conducted in each plot for a total of six replications per treatment.

332

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

2.4. Statistical analysis Differences in aggregate size distribution, soil mean weight diameter, aggregate-associated C, soil bulk density, near-saturated hydraulic conductivity, and saturated hydraulic conductivity among soil treatments at the same depth in the same year were determined by oneway analysis of variance using PROC GLM in SAS software (SAS Institute, Inc., Cary, NC) and Tukey's HSD at α = 0.05. Differences of aggregate size distribution at the same soil depth between 2011 and 2012 were determined by t-test analysis using PROC TTEST in SAS software (SAS Institute, Inc., Cary, NC).

Table 2 Soil mean weight diameter (MWD) of soils from typical urban land development practice (TP), typical practice enhanced by tilling (ET), a post-development soil rehabilitation practice (PR), and an undisturbed control (UN) at 0–5 cm, 5–10 cm, and 15–30 cm soil depths four and five years after treatment. Year

Soil depth

Treatments TP

ET

PR

UN

0.68 (0.06)b 0.82 (0.11) 0.67 (0.09) 0.89 (0.47) 0.76 (0.17) 0.66 (0.17)

0.68 (0.10)b 0.83 (0.10) 0.68 (0.03) 0.72 (0.09) 0.77 (0.18) 0.70 (0.17)

0.90 (0.10)a 0.99 (0.16) 0.69 (0.07) 0.91 (0.15) 0.96 (0.17) 0.92 (0.55)

MWDa 2011

2012

3. Results

0–5 cm 5–10 cm 15–30 cm 0–5 cm 5–10 cm 15–30 cm

0.74 (0.08)bb 0.89 (0.15) 0.62 (0.09) 0.76 (0.16) 0.87 (0.14) 0.70 (0.17)

a

Values in parentheses represent standard errors of the means (n = 6). Within the same soil depth and year, letters denote differences among treatments using Tukey's HSD at α = 0.05. No letters indicates no significant differences at α = 0.05.

3.1. Aggregate size distribution

b

In 2011, four years after simulated land urbanization and soil rehabilitation, UN had a greater proportion of macroaggregates, but smaller proportions of microaggregates and fine fraction than the other three treatments in the top 10 cm of soil, except for the fine fraction at the 5–10 cm depth (Table 1). However, at 15–30 cm soil depth, there was only weak evidence of differences in aggregate size distribution across all treatments (p-values = 0.06, 0.20 and 0.10, for macroaggregates, microaggregates, and fine fraction respectively) (Table 1). At this depth, UN had slightly higher macroaggregates (0.26 g sand-free aggregate g−1 soil) than the three other treatments (ranging from 0.20 to 0.22 g sand-free aggregate g−1 soil). Although statistical evidence was inconclusive that this difference was treatment related, the same pattern was observed in 2012 (overall p-value = 0.20). In 2012, soil aggregate size distribution exhibited similar trends among treatments as in the previous year (Table 1). In the top 10 cm of soil, treatments subjected to urban land development followed by post-development practices (i.e., ET, TP, and PR) had lower macroaggregate proportions but greater microaggregate and fine fraction proportions compared to UN soils. However, as in 2011, treatment effect on aggregate size classes at 15–30 cm soil depth was inconclusive. There was no evidence showing that soil MWD was affected by different soil treatments at all three soils depths in both 2011 and 2012 (ranging from 0.62 to 0.99) except 0–5 cm soil depth in 2011 when UN had

higher MWD than other three treatments (Table 2). However, at 0–5 cm soil depth, the MWD of the other three soil treatments (TP, ET, and PR) was closer to UN in 2012 than in 2011, suggesting that aggregation is re-developing in this study system. Overall, when these three soil treatments (TP, ET, and PR) were pooled, the proportion of macroaggregates increased from 2011 to 2012 at 0–5 cm and 15–30 cm depths (F = 3.75 and 3.28; p b 0.01 and = 0.02; respectively).

3.2. Aggregate-associated carbon In 2011, at 0–5 cm, UN had the greatest macroaggregate and microaggregate C concentrations and C pool sizes; other than greater microaggregate C concentrations these effects were no longer evident at 5–10 cm. In 2012, macroaggregate C concentrations in post-development treatments were greater than in 2011 at 0– 10 cm, particularly in PR, and thus differences between disturbed and undisturbed treatments were no longer evident (Fig. 1, Table 3). Nonetheless, at 0–5 cm depth, UN had greater aggregate C pool sizes than the other three treatments (Table 3). In contrast, in 2011 at 15–30 cm, PR had the greatest macroaggregate and

Table 1 Macroaggregates (250–2000 μm), microaggregates (53–250 μm), and fine fraction (b53 μm) of soils of typical urban land development practice (TP), typical practice enhanced by tilling (ET), a post-development soil rehabilitation practice (PR), and an undisturbed control (UN) at 0–5 cm, 5–10 cm, and 15–30 cm soil depths four and five years after treatment. Year

Soil depth

Aggregates

Treatments TP

ET

PR

UN

0.22 (0.02)b 0.19 (0.02)a 0.18 (0.02) a 0.37 (0.04)b 0.16 (0.03)ab 0.10 (0.01) 0.20 (0.02) 0.29 (0.03) 0.13 (0.02) 0.30 (0.04)b 0.19 (0.02)a 0.20 (0.02)ab 0.35 (0.04)b 0.21 (0.03)a 0.15 (0.03)a 0.29 (0.03) 0.26 (0.02) 0.17 (0.02)

0.34 (0.02)a 0.13 (0.01)b 0.12 (0.01)b 0.44 (0.02)a 0.12 (0.02)b 0.08 (0.02) 0.26(0.04) 0.26 (0.03) 0.11(0.02) 0.43 (0.02)a 0.13 (0.01)b 0.16 (0.02)b 0.48 (0.03)a 0.14 (0.02)b 0.11 (0.02)b 0.38 (0.09) 0.21 (0.05) 0.12 (0.04)

g sand-free aggregate per g soila 2011

0–5 cm

5–10 cm

15–30 cm

2012

0–5 cm

5–10 cm

15–30 cm

a

Macroaggregate Microaggregate Fine fraction Macroaggregate Microaggregate Fine fraction Macroaggregate Microaggregate Fine fraction Macroaggregate Microaggregate Fine fraction Macroaggregate Microaggregate Fine fraction Macroaggregate Microaggregate Fine fraction

0.23 (0.02)bb 0.18(0.02)a 0.18 (0.02)a 0.35 (0.04)b 0.18 (0.04)a 0.10 (0.03) 0.22 (0.04) 0.30 (0.04) 0.12 (0.03) 0.34 (0.04)b 0.16 (0.03)ab 0.21 (0.02)a 0.27 (0.04)c 0.18 (0.03)ab 0.14 (0.02)ab 0.34 (0.08) 0.27 (0.05) 0.14 (0.03)

0.23 (0.03)b 0.20 (0.04)a 0.16 (0.02)a 0.37 (0.06)ab 0.18 (0.05)a 0.10 (0.02) 0.22 (0.04) 0.30 (0.04) 0.11 (0.02) 0.29 (0.07)b 0.18 (0.04)a 0.19 (0.04)ab 0.35 (0.06)b 0.21 (0.04)a 0.16 (0.02)a 0.33 (0.08) 0.27 (0.06) 0.13 (0.04)

Values in parentheses represent standard errors of the means (n = 6). Within the same aggregate size, soil depth, and year, letters denote differences among treatments using Tukey's HSD at α = 0.05. No letters indicates no significant differences at α = 0.05. b

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

333

Fig. 1. Carbon concentration in macroaggregate (250–2000 μm) and microaggregate (53–250 μm) at three soil depths of typical urban land development practice (TP), typical practice enhanced by tilling (ET), a post-development soil rehabilitation practice (PR), and an undisturbed control (UN) in 2011 (A) and 2012 (B). Error bars depict standard errors of the means (n = 6). Within each size aggregate at the same soil depth, letters denote differences among treatments using Tukey's HSD at α = 0.05. NS indicates no significant differences at α = 0.05.

microaggregate C concentrations and macroaggregate C pool sizes among all treatments, including UN (Fig. 1, Table 3). In 2012, however, PR only increased macroaggregate C concentration and pool size, but had no effect on microaggregate C concentration and pool size (Fig. 1 and Table 3). Thus, PR increased macroaggregateassociated C concentration and C pool size in four and five years after installation.

3.3. Soil bulk density Soil bulk density was not affected by soil treatment at 2.5–7.6 cm depth. At 15.2–20.3 cm depth, however, PR had a significantly lower bulk density (1.49 g cm− 3) than other treatments (treatment means ranging from 1.61 to 1.76 g cm− 3). In deeper soil regions, (30.5–35.6 cm and 50.8–55.9 cm), all bulk densities were relatively

Table 3 Carbon pool size of macroaggregates (250–2000 μm) and microaggregates (53–250 μm) of typical urban land development practice (TP), typical practice enhanced by tilling (ET), a postdevelopment soil rehabilitation practice (PR), and an undisturbed control (UN) soils at 0–5 cm, 5–10 cm, and 15–30 cm soil depths four and five years after treatment. Year

Soil depth

Aggregates

Treatments TP

ET

PR

UN

8.24 (2.10)b 8.51 (1.14)b 10.09 (4.41) 8.91 (1.54) 9.30 (4.64)a 6.89 (3.60) 8.69 (2.95)ab 9.29 (1.87)b 10.37 (4.19) 10.30 (2.60) 7.72 (4.62)a 7.08 (5.91)

14.35 (2.25)a 16.15 (3.16)a 9.52 (1.28) 11.17 (2.28) 4.38 (1.81)b 5.71 (1.63) 9.77 (1.25)a 13.19 (1.53)a 8.96 (1.10) 11.22 (2.04) 5.20 (1.15)ab 5.76 (1.82)

−1 a

Carbon pool size (Mg C ha 2011

0–5 cm 5–10 cm 15–30 cm

2012

0–5 cm 5–10 cm 15–30 cm

a

Macroaggregate Microaggregate Macroaggregate Microaggregate Macroaggregate Microaggregate Macroaggregate Microaggregate Macroaggregate Microaggregate Macroaggregate Microaggregate

7.60 (1.70)bb 9.48 (1.92)b 8.22 (1.76) 9.72 (2.01) 2.66 (0.61)b 3.46 (0.69) 7.55 (2.49)ab 9.56 (2.53)b 9.57 (2.91) 10.96 (2.62) 2.94 (0.91)b 3.49 (0.81)

)

6.34 (1.86)b 8.59 (2.11)b 7.61 (1.65) 8.98 (2.08) 2.74 (1.77)b 3.71 (2.44) 6.03 (1.83)b 8.29 (2.22)b 6.93 (1.95) 9.08 (2.46) 2.85 (1.95)b 3.74 (2.19)

Values in parentheses represent standard errors of the means (n = 6). Within the same aggregate size, soil depth, and year, letters denote differences among treatments using Tukey's HSD at α = 0.05. No letters indicates no significant differences at α = 0.05. b

334

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

high (means ranging from 1.72 to 1.79 g cm−3 and 1.67 to 1.76 g cm−3, respectively) although bulk density in PR plots at 50.8–55.9 cm was significantly greater than in ET and TP at α = 0.05. 3.4. Soil hydraulic conductivity Surface Knear was not affected by soil treatment (ranging from 0.4 to 2.3 cm h− 1) (Fig. 2). Lower in the soil profile, however, there were pronounced differences in Ksat among treatments. At 10–25 cm and 25–40 cm soil depths, Ksat was twice as rapid in PR soils (14.8 and 6.3 cm h − 1 , respectively) compared to UN soils (6.7 and 3.0 cm h − 1 , respectively) and as much as 10 times more rapid than in the TP and ET treatments (ranging from 1.3 to 1.7 cm hr−1 and ranging from 1.0 to 3.7 cm h−1, respectively) (Fig. 2). 4. Discussion 4.1. Impact of urban land development and post-development soil rehabilitation on soil structure and function Physical disturbance from urban land development and postdevelopment practices appear to have crushed macroaggregates leading to increased microaggregate and fine fractions in surface soils. This finding is consistent with previous studies following other types of disturbance with varying levels of severity, such as tillage in agricultural and reclamation following surface mining

Fig. 2. Mean unsaturated hydraulic conductivity at surface when pressure head h = −2.0 cm (n = 18) and saturated hydraulic conductivity (Ksat) at 25 cm and 40 cm soil depths (n = 6) of typical urban land development practice (TP), typical practice enhanced by tilling (ET), a post-development soil rehabilitation practice (PR), and an undisturbed control (UN) five years after treatment installation. Values in parentheses represent standard error of the mean. Within each soil depth, superscript letters indicate significant differences across soil treatments at α = 0.05 using Tukey's HSD. NS indicates no significant differences at α = 0.05.

(Lawal et al., 2009; Six et al., 2000, 2004; Wick et al., 2009b). Soil disturbance from tillage has been recognized as a major cause of reduction in the number and stability of soil aggregates when land changes from natural/grassland settings into agricultural use (Six et al., 2000). Soil manipulations during urbanization partly mimic this process. Urban areas generally experience a different set of human-induced disturbances compared to agricultural land. For example, urban land development processes may involve vegetation clearance, topsoil removal, stockpiling, compaction, building, or soil replacement all of which can drastically affect soil aggregation (Wick et al., 2009b). In our study, topsoil removal and replacement, for example, resulted in about 29% average reduction in the proportion of macroaggregates in the surface soils even without tillage. Our finding of C loss in surface soils after disturbance is consistent with other studies (Mikha and Rice, 2004; Six et al., 1998; Wick et al., 2009b). The increase in microaggregates and fine fraction at the expense of macroaggregates suggest that this C loss is due to disturbance breaking aggregates apart, leading to organic matter exposure and thus easier access for heterotrophic soil microbes (Rovira and Greacen, 1957). In the present study, soil scraping, stockpiling, and replacement had a profound effect on surface soil structure although, notably, additional disturbance via tillage (as in the ET treatment) did not lead to further degradation of soil aggregation. On the other hand, compaction alone, as is indicated by the similar aggregate size distribution among treatments at 15–30 cm soil depth, had little effect on aggregate size distribution after four years. Interestingly, the PR treatment also had similar aggregate size distributions at 15–30 cm, suggesting that either (A) the backhoe-style tillage (that separates soil into large clumps from approximately 5–30 cm in diameter) combined with compost addition did not disrupt aggregates as much as might be expected, or (B) that aggregates had reformed within the study time period. As might be expected, surface soils not only are more vulnerable to disturbance during manipulation processes (Six et al., 1998), but also show the fastest recovery post-disturbance (Grandy and Robertson, 2007). Thus, efforts directed at redeveloping subsurface soils have potential to provide relatively quick recovery and then persist over time, even if surface soils are subjected to additional disturbances such as might occur during urban land use. Restoring degraded subsoils may provide opportunities for additional strategies for improving soil functions for the long-term. We did not find evidence, however, that PR improved aggregation five years after installation compared to other treatments, although macroaggregate C concentration was considerably greater in rehabilitated subsoils. Our results differ from other studies that observed rapid changes in aggregation after compost application (Tejada et al., 2009; Whalen et al., 2003). However, these studies focused on aggregate stability (Tejada et al., 2009) or observed changes in very large aggregates (N4 mm) (Whalen et al., 2003) rather than macro- and microaggregate distribution as defined in the present study. We found that compost amendment with deep tillage increased macroaggregate-associated C concentration and pool size, but also led to more stable microaggregate-associated C four years after installation. Although C inputs increased macroaggregate-associated C concentration and C pool size during our five-year study, repairing disrupted aggregation, which is important for enhanced long-term C storage as well as soil hydrologic properties, did not occur.

4.2. Soil hydraulic conductivity change resulting from typical urban land development In our study, surface Knear was not affected by treatment (Fig. 2). Surface Knear was low (0.44 to 2.28 cm h−1), likely because there was no surface cover treatment (e.g., grass, mulch). However, low conductivity was not due to potential surface crusting, because measurements with approximately 3 mm surface soil removed were similar. In our

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

system, surface soils were the limiting layer for water movement through the soil profile for both UN and PR. Past studies have found that infiltration rates decrease after urbanization (Gregory et al., 2006; Pitt et al., 2008; Yang and Zhang, 2011). In our study, the mean surface Knear of TP soil (1.1 cm h−1) was near the lower bound values reported in other similar studies (Hamilton and Waddington, 1999; Woltemade, 2010) (Fig. 2). However, unlike these studies which used double ring infiltrometers, our measurements were made under slight tension, thereby reducing the influence of very large macropores or other preferential flow paths (Beven and Germann, 2013). Most studies have focused on surface soils, yet restricting layers may occur deeper in the profile. At these greater soil depths treatment differences were pronounced in our study. At 10–40 cm, for example, postdevelopment rehabilitated soils (PR) Ksat averaged 10.6 cm h−1 approximately twice that of UN soils (4.9 cm h−1) and much greater than the 1.2 cm h−1 average of TP soils (Fig. 2). Reduced Ksat in TP might be due to macropore clogging by microaggregates and the fine soil fraction (Richard et al., 2001). Although subsurface manipulations did not increase aggregate formation, they had a profound effect on Ksat. Since Ksat can be dominated by macropores and cracks, these manipulations might increase macropores and connectivity of large pores (Kodešová et al., 2011) thus creating favorable flow paths for water in subsurface soils (Beven and Germann, 2013). Increases in Ksat might also be indirectly related to documented increased soil C concentration and content (Chen et al., 2013) since those could be indications that soil aggregate formation is in progress, even though no measurable aggregate formation occurred during the 5-year study period. In a study using a double-ring infiltrometer, Olson et al. (2013) also found that compost amendment increased Ksat, however, to a smaller degree than was found in our study (2.7– 5.7 times that of the control). The increase resulting from PR in our study may be partly influenced by the presence of trees. Although all treatments included tree planting, tree root distribution likely varies in response to soil physical characteristics (Day et al., 2010). Tree roots can improve hydraulic conductivity significantly by deep root penetration (Bartens et al., 2008; Johnson and Lehmann, 2006) and soil structure improvement (Bottinelli et al., 2013). The increases in Ksat resulting from PR at 10–25 and 25–40 cm depths are significant and suggest that soil management has potential to play an important role in urban stormwater runoff abatement. Assignments to a hydrologic soil group (HSG) that corresponds to a minimum hydraulic conductivity for a given soil or other estimates of soil hydraulic conductivity are frequently made to facilitate runoff estimation. Although on average soils may drain more poorly after urban land use change, our study suggests that conductivity can be altered by soil management to a degree that would influence runoff estimates. 4.3. Relation between water movement and soil structure We found that disturbance resulting from urban land development broke macroaggregates into smaller aggregates in surface soils and reduced Ksat in subsurface soils; but had only a slight effect on aggregate-associated C concentration and pool size. As we expected, PR resulted in reduced soil bulk density and increased aggregateassociated C concentrations and Ksat in subsurface soils after five years. Although there was no aggregate size distribution difference detected at 15–30 cm among treatments in either year, aggregation was improved in PR from 2011 to 2012 which showed higher proportions of both macroaggregates and microaggregates (F = 3.73 and 2.57; p b 0.01 and = 0.06; respectively). Furthermore, soil bulk density was considerably lower in this region in PR plots compared to TP (1.49 g cm − 3 vs. 1.76 g cm− 3). Nonetheless, more than five years may be needed to observe clear differences in aggregation among treatments. In spite of these indications that accelerated aggregation may be underway in PR plots and contributing to observed increases

335

in Ksat at 10–40 cm soil depth, other factors are likely influencing Ksat. Jirků et al. (2013), for example, could not establish correlations between soil structure and aggregate stability and soil hydraulic properties in agricultural topsoils, and suggested that this may be due to high variability in hydraulic properties. Increased Ksat in the PR treatment might result from more macropores and better connectivity after rehabilitation (Pagliai et al., 2004) but at a scale not described by aggregate size distribution analysis. Our study investigated both surface hydraulic conductivity and subsurface water flow which suggest that PR had a significant impact on subsurface water movement and storage. Thus, urban hydrology could be strongly influenced by soil management practices and typical measures of soil structure may not be adequate to assess these changes. Based on our findings and those from other studies, K sat in urban systems is related to a combination of soil physical (aggregation, macro-porosity/cracking, pore clogging and discontinuity and compaction) and biological properties (organic matter and rooting channels). In other words, the change of water flow in soils cannot be explained or estimated by an individual index as originally anticipated, and larger scale physical discontinuities in the soil profile must be considered. 5. Conclusions Typical urban land development practices disrupted soil aggregation, reducing the proportion of macroaggregates and leading to macroaggregate-associated C loss in surface soils. Compacted urban soils resulting from typical urban land development practices decreased soil saturated hydraulic conductivity compared to pre-development undisturbed soils. However, post-development soil rehabilitation via a deep tillage-compost amendment-tree planting system improved soil hydraulic conductivity significantly at lower depths where compacted soil was broken into large clods with compost addition even though there was no evidence that this deep tillage affected subsurface soil aggregate size distribution. Thus, the protection of soils at the early stage of urban development is critical to avoid significant soil structure damage, C loss and increases in stormwater runoff. Higher soil hydraulic conductivity in subsurface soils due to profile rebuilding indicates that post-development rehabilitation can improve soil physical characteristics that affect water movement (i.e. macro-pores and cracks). The ability of urban soil management to affect its hydrologic soil group should be considered in planning and urban hydrologic transport models, especially when assessing the impact of site management practices. Acknowledgments This research is funded in part by the Tree Research and Education Endowment Fund 08-HJ-06 and the Institute for Critical Technology and Applied Science at Virginia Tech, the Virginia Agricultural Experiment Station and the McIntire Stennis program of the National Institute of Food and Agriculture U.S. Department of Agriculture, and the Garden Club of America's urban forestry fellowship program. The authors gratefully acknowledge Rachel Layman, Velva Groover, and J. Roger Harris for their assistance and support. Graphical abstract prepared by Sarah Gugercin. References Abu-Sharar T, Bingham F, Rhoades J. Reduction in hydraulic conductivity in relation to clay dispersion and disaggregation. Soil Sci Soc Am J 1987;51:342–6. Alaoui A, Lipiec J, Gerke H. A review of the changes in the soil pore system due to soil deformation: a hydrodynamic perspective. Soil Tillage Res 2011;115–116:1–15. Bartens J, Day SD, Harris JR, Dove JE, Wynn TM. Can urban tree roots improve infiltration through compacted subsoils for stormwater management? J Environ Qual 2008;37: 2048–57. Berthier E, Andrieu H, Creutin J. The role of soil in the generation of urban runoff: development and evaluation of a 2D model. J Hydrol 2004;299:252–66. Beven K, Germann P. Macropores and water flow in soils revisited. Water Resour Res 2013;49:3071–92.

336

Y. Chen et al. / Science of the Total Environment 494–495 (2014) 329–336

Bottinelli N, Menasseri‐Aubry S, Cluzeau D, Hallaire V. Response of soil structure and hydraulic conductivity to reduced tillage and animal manure in a temperate loamy soil. Soil Use Manag 2013;29:401–9. Boyle M, Frankenberger W, Stolzy L. The influence of organic matter on soil aggregation and water infiltration. J Prod Agric 1989;2:290–9. Brown S, Cotton M. Changes in soil properties and carbon content following compost application: results of on-farm sampling. Compost Sci Util 2011;19:87–96. Carminati A, Flühler H. Water infiltration and redistribution in soil aggregate packings. Vadose Zone J 2009;8:150–7. Chen Y, Day SD, Wick AF, Strahm BD, Wiseman PE, Daniels WL. Changes in soil carbon pools and microbial biomass from urban land development and subsequent postdevelopment soil rehabilitation. Soil Biol Biochem 2013;66:38–44. Cogger CG. Potential compost benefits for restoration of soils disturbed by urban development. Compost Sci Util 2005;13:243–51. Collins KA, Lawrence TJ, Stander EK, Jontos RJ, Kaushal SS, Newcomer TA, et al. Opportunities and challenges for managing nitrogen in urban stormwater: a review and synthesis. Ecol Eng 2010;36:1507–19. Day SD, Wiseman PE, Dickinson SB, Harris JR. Contemporary concepts of root system architecture of urban trees. Arboricult Urban For 2010;36:149–59. De Lucia B, Cristiano G, Vecchietti L, Bruno L. Effect of different rates of composted organic amendment on urban soil properties, growth and nutrient status of three Mediterranean native hedge species. Urban For Urban Green 2013;12:537–45. Denef K, Six J, Bossuyt H, Frey SD, Elliott ET, Merckx R, et al. Influence of dry–wet cycles on the interrelationship between aggregate, particulate organic matter, and microbial community dynamics. Soil Biol Biochem 2001;33:1599–611. Elliott E, Palm C, Reuss D, Monz C. Organic matter contained in soil aggregates from a tropical chronosequence: correction for sand and light fraction. Agr Ecosyst Environ 1991;34:443–51. Grandy AS, Robertson GP. Land-use intensity effects on soil organic carbon accumulation rates and mechanisms. Ecosystems 2007;10:59–74. Gregory JH, Dukes MD, Jones PH, Miller GL. Effect of urban soil compaction on infiltration rate. J Soil Water Conserv 2006;61:117–24. Hamilton G, Waddington D. Infiltration rates on residential lawns in central Pennsylvania. J Soil Water Conserv 1999;54:564–8. Jim CY. Soil compaction as a constraint to tree growth in tropical & subtropical urban habitats. Environ Conserv 1993;20:35–49. Jim CY. Physical and chemical properties of a Hong Kong roadside soil in relation to urban tree growth. Urban Econ 1998a;2:171–81. Jim CY. Urban soil characteristics and limitations for landscape planting in Hong Kong. Landsc Urban Plan 1998b;40:235–49. Jirků V, Kodešová R, Nikodem A, Mühlhanselová M, Žigová A. Temporal variability of structure and hydraulic properties of topsoil of three soil types. Geoderma 2013; 204:43–58. Johnson MS, Lehmann J. Double-funneling of trees: stemflow and root-induced preferential flow. Ecoscience 2006;13:324–33. Kay B. Soil structure and organic carbon: a review. In: Lal R, Kimble JM, Follett RF, Steward BA, editors. Soil processes and the carbon cycle. Boca Raton, FL: CRC Press; 1998. p. 169–97. Khaleel R, Reddy K, Overcash M. Changes in soil physical properties due to organic waste applications: a review. J Environ Qual 1981;10:133–41. Kodešová R, Jirků V, Kodeš V, Mühlhanselová M, Nikodem A, Žigová A. Soil structure and soil hydraulic properties of Haplic Luvisol used as arable land and grassland. Soil Tillage Res 2011;111:154–61. Lawal HF, Ogunwole JO, Uyovbisere EO. Changes in soil aggregate stability and carbon sequestration mediated by land use practices in a degraded dry savanna Alfisol. Trop Subtrop Agroecosyst 2009;10:423–9. Layman RM. Rehabilitation of severely compacted urban soil to improve tree establishment and growth. [Master's thesis] Horticulture. Master of Science Virginia Polytechnic Institute and State University; 2010. Leopold LB. Hydrology for urban land planning: a guidebook on the hydrologic effects of urban land use. U.S. Geological Survey Circular, 554. United States: Department of the Interior; 1968. [554].

Lipiec J, Wójciga A, Horn R. Hydraulic properties of soil aggregates as influenced by compaction. Soil Tillage Res 2009;103:170–7. Martens D, Frankenberger W. Modification of infiltration rates in an organic-amended irrigated. Agron J 1992;84:707–17. Mikha MM, Rice CW. Tillage and manure effects on soil and aggregate-associated carbon and nitrogen. Soil Sci Soc Am J 2004;68:809–16. Milesi C, Elvidge CD, Nemani RR, Running SW. Assessing the impact of urban land development on net primary productivity in the southeastern United States. Remote Sens Environ 2003;86:401–10. Olson NC, Gulliver JS, Nieber JL, Kayhanian M. Remediation to improve infiltration into compact soils. J Environ Manage 2013;117:85–95. Pagliai M, Vignozzi N, Pellegrini S. Soil structure and the effect of management practices. Soil Tillage Res 2004;79:131–43. Paul MJ, Meyer JL. Streams in the urban landscape. Annu Rev Ecol Evol Syst 2001:333–65. Pitt R, Clark SE. Integrated storm–water management for watershed sustainability. J Irrig Drain Eng 2008;134:548–55. Pitt R, Lantrip J, Harrison R, Henry CL, Xue D, O'Connor T. Infiltration through disturbed urban soils and compost-amended soil effects on runoff quality and quantity. Washington, D.C.: U.S. Environmental Protection Agency, Office of Research and Development; 1999 [No. EPA/600/R-00/016]. Pitt R, Chen S-E, Clark SE, Swenson J, Ong CK. Compaction's impacts on urban storm– water infiltration. J Irrig Drain Eng 2008;134:652–8. Rawls W, Pachepsky YA, Ritchie J, Sobecki T, Bloodworth H. Effect of soil organic carbon on soil water retention. Geoderma 2003;116:61–76. Richard G, Cousin I, Sillon J, Bruand A, Guérif J. Effect of compaction on the porosity of a silty soil: influence on unsaturated hydraulic properties. Eur J Soil Sci 2001;52:49–58. Rovira A, Greacen E. The effect of aggregate disruption on the activity of microorganisms in the soil. Crop Pasture Sci 1957;8:659–73. Sanders RA. Urban vegetation impacts on the hydrology of Dayton, Ohio. Urban Ecol 1986;9:361–76. Seto KC, Güneralp B, Hutyra LR. Global forecasts of urban expansion to 2030 and direct impacts on biodiversity and carbon pools. Proc Natl Acad Sci 2012;109:16083–8. Six J, Elliott ET, Paustian K, Doran JW. Aggregation and soil organic matter accumulation in cultivated and native grassland soils. Soil Sci Soc Am J 1998;62:1367–77. Six J, Elliott E, Paustian K. Soil macroaggregate turnover and microaggregate formation: a mechanism for C sequestration under no-tillage agriculture. Soil Biol Biochem 2000; 32:2099–103. Six J, Bossuyt H, Degryze S, Denef K. A history of research on the link between (micro)aggregates, soil biota, and soil organic matter dynamics. Soil Tillage Res 2004;79:7–31. Sloan JJ, Ampim PAY, Basta NT, Scott R. Addressing the need for soil blends and amendments for the highly modified urban landscape. Soil Sci Soc Am J 2012;76:1133–41. Tejada M, Hernandez M, Garcia C. Soil restoration using composted plant residues: effects on soil properties. Soil Tillage Res 2009;102:109–17. Tisdall J, Oades J. Organic matter and water-stable aggregates in soils. J Soil Sci 1982;33: 141–63. Whalen JK, Hu Q, Liu A. Compost applications increase water-stable aggregates in conventional and no-tillage systems. Soil Sci Soc Am J 2003;67:1842–7. Wick AF, Ingram LJ, Stahl PD. Aggregate and organic matter dynamics in reclaimed soils as indicated by stable carbon isotopes. Soil Biol Biochem 2009a;41:201–9. Wick AF, Stahl PD, Ingram LJ, Vicklund L. Soil aggregation and organic carbon in shortterm stockpiles. Soil Use Manag 2009b;25:311–9. Woltemade CJ. Impact of residential soil disturbance on infiltration rate and stormwater runoff. J Am Water Resour Assoc 2010;46:700–11. Xiao Q, McPherson EG. Performance of engineered soil and trees in a parking lot bioswale. Urban Water J 2011;8:241–53. Xiao Q, McPherson EG, Ustin SL, Grismer ME, Simpson JR. Winter rainfall interception by two mature open-grown trees in Davis, California. Hydrol Process 2000;14:763–84. Yang J-L, Zhang G-L. Water infiltration in urban soils and its effects on the quantity and quality of runoff. J Soils Sediment 2011;11:751–61.