Soil Carbon and Nitrogen Dynamics beneath ...

Published online April 19, 2018

Soil & Water Management & Conservation

Soil Carbon and Nitrogen Dynamics beneath Impervious Surfaces Hamed Majidzadeh*

School of Forestry and Wildlife Sciences Auburn Univ. Auburn, AL 36849 and Biogeochemistry & Environ. Quality Lab. Clemson Univ. Clemson, SC 29634

B. Graeme Lockaby Robert Price Robin Governo

School of Forestry and Wildlife Sciences Auburn Univ. Auburn, AL 36849

Soil sealing by impervious surfaces is a major disturbance caused by urbanization and has been shown to reduce soil carbon and nitrogen significantly. However, the degree to which these changes are driven by the initial disturbance (i.e., top soil removal) or post-construction processes is not clear. A controlled field study consisting of three treatments including concrete slab (SLB), home with crawl space (CRW), and control (CNT) was conducted to monitor changes in soil properties immediately after sealing and over a 15-mo time frame. At depth 10 to 20 cm (first layer beneath the concrete slab) soil carbon decreased by 30.4% (±3.4%) from 2.3 (±0.4) kg m–2 at Month 0 to 1.57 (±0.36) kg m–2 at the end of experiment with a rate of 0.04 (±0.01) kg m–2 per month (p = 0.001, F1,7 = 9.27). At this depth soil, carbon and nitrogen fluctuated seasonally at CRW and CNT plots. At depth 20 to 30 cm, the soil carbon beneath the CRW (1.14 ± 0.21 kg m–2) was higher than that beneath the SLB (0.93 ± 0.23 kg m–2, p = 0.01, F2,23 = 5.26), suggesting that partially permeable impervious surfaces such as CRW can support limited inputs of carbon and organic matter. In addition to carbon and nitrogen, temporal changes of microbial biomass carbon and nitrogen mineralization rates were also monitored. Nitrogen mineralization rates also decreased below impervious surfaces as evidenced by zero net ammonium production rate. However, a significant increase in mineralization rates was observed in warm periods of the year beneath SLB treatments.

U

Core Ideas • Carbon loss from soil beneath impervious surfaces would not exceed 1.92 Pg globally. • Net nitrogen mineralization rates beneath impervious surfaces were primarily related to nitrification. • Temporally, nitrate accumulation was evident beneath impervious surfaces. • Soil microbial biomass carbon was the key factor affecting soil carbon beneath the concrete slabs. • Soil carbon beneath homes was positively related to soil volumetric water content. Soil Science Society of America Journal

rbanization alters soil physical and chemical properties including soil carbon storage and sequestration rates (Scharenbroch et al., 2005; Pouyat et al., 2006). To evaluate the effects of urbanization on the ecosystem, and the global carbon cycle, it is important to quantify these changes in various urban land uses and land cover-types (Lal, 2004; Yan et al., 2015). Soil sealing by impervious surfaces such as buildings, roads, sidewalks, and parking lots is a major disturbance affecting soil physical and chemical properties by limiting water and gas exchange and restricting the input of organic matter (Montague and Kjelgren, 2004; Lorenz and Lal, 2009). Moreover, the construction process usually involves compaction and removal of organic-rich top soil (Scalenghe and Marsan, 2009; Yan et al., 2015). Besides soil biogeochemical properties, impervious surfaces and in general soils of urban, industrial, traffic, mining, and military areas (SUITMAs) reduce the infiltration rate and as a result increase the surface runoff and enhance the risk of flooding (Morel et al., 2015). Also, they contribute to the urban heat island effect due to their lower albedo compared with rural and agricultural lands (Morel et al., 2015; Zhu et al., 2017). Impervious surfaces cover more than one-third of urban areas in the United States and exceed 130,000 km2 (Elvidge et al., 2004). Despite the importance and magnitude of changes caused by impervious surfaces, few studies have sampled soils beneath them (Raciti et al., 2012; Wei et al., 2014a; Piotrowska-Długosz and Charzyński, 2015; Yan et al., 2015; Mendyk and Charzyński, 2016). These studies have reported a 68 to 75% decrease in soil carbon underneath sidewalks, residential Soil Sci. Soc. Am. J. doi:10.2136/sssaj2017.11.0381 Received 6 Nov. 2017. Accepted 12 Mar. 2018. *Corresponding author ([email protected]). © Soil Science Society of America, 5585 Guilford Rd., Madison WI 53711 USA. All Rights reserved.

pavements, and residential driveways compared with open areas. A significant decline in soil nitrogen (Raciti et al., 2012; Wei et al., 2014a) and microbial activity (Wei et al., 2014b; PiotrowskaDługosz and Charzyński, 2015) have also been reported. However, it is not clear if these changes have occurred over time or during initial construction disturbances. We do not know if the soil beneath impervious surfaces become biologically inactive or if the soil serves as a carbon sink or source after disturbance and sealing. The inputs of organic matter beneath impervious surfaces are very limited but may occur to some extent through root mortality, and insect activity (Lehmann and Stahr, 2007; Scalenghe and Marsan, 2009). Carbon outputs are also limited in these systems because of the significant decrease in oxygen, microbial activity, and leaching (Wei et al., 2013). Moreover, sizes and types of impervious surfaces may have different effects on soil properties. Yan et al. (2015) showed that the size of impervious surfaces is a factor of influence and that soil carbon decreases horizontally from the edge of the pavement to the center. Homes and concrete slabs are two major types of impervious surfaces. Homes account for more than 20% of impervious surfaces in the United States (US Census Bureau, 2013). Multiunit homes and tall building construction involve intensive disturbances as well as major top soil removal and is assumed to have zero carbon storage (Yan et al., 2015). However, only 23% of the homes in the United States are multi-units. Approximately 53% of the houses in the United States are manufactured homes (6%), homes built on crawl space (16%), and homes constructed with a basement (31%) (US Census Bureau, 2013). Crawl-space foundations are to elevate the homes off the floor by around 40 to 100 cm using concrete blocks or bricks. To examine the impacts of soil sealing on soil physical and biological properties, a controlled field study consisting of three treatments was conducted, including concrete slabs (SLB), structures simulating houses built on a crawl space (CRW), and grassed control (CNT) plots. Soil sealing has been classified into two levels of total imperviousness (i.e., solid concrete and asphalt) and partial permeability which allows some penetration of moisture

and air (Charzyński et al., 2017). In this study, the SLB represented total importability, and the CRW treatment is representative of the partial permeable impervious surfaces. Soil physical, chemical, and biological properties were monitored on these plots for a year. The objectives of this study were to (i) monitor changes in physical properties such as temperature and moisture over time; (ii) evaluate changes in soil carbon and nitrogen storage and dynamics beneath impervious surfaces; (iii) quantify biological properties such as microbial biomass, nitrogen mineralization, and nitrification beneath impervious surfaces over time.

MaterialS and methods Site Construction

The study site was on an unfertilized field covered with grasses near Auburn, AL (latitude 32.60° N, longitude 85.48° W), which is located on the fall line between the Coastal Plain and the Piedmont Plateau. Auburn has a mean annual air temperature of 17.4°C and mean precipitation of 1337 mm. Soil on the site was classified as a fine, kaolinitic, thermic Rhodic Kanhapludults (Acrisol according to World Reference Base for Soil Resources [IUSS Working Group WRB, 2015]). Treatments were installed using 5-m by 5-m plots and replicated four times based on randomized complete block design (RCBD), totaling twelve plots (Fig. 1A). Vegetation and the top 10 cm of soil were removed for construction of SLB and CRW plots. The soil was compacted by tamping prior to the installation of concrete slabs to mimic normal construction practices. While the tamping resulted in soil compaction underneath the concrete, the concrete slab treatment would not have been realistic without the inclusion of the tamping activity. Afterward, a 10-cm thick concrete slab was poured on October 2014. Eighteen blocks were situated in the middle of each concrete slab for individual removal during sampling (Fig. 1B). The cracks between blocks were sealed using silicon after each sampling to prevent moisture infiltration. Local contractors were consulted in the design of the CRW treatment which consisted of a 60 cm (H) by 5 m (L) by

Fig. 1. (A) Three treatments with four replicates totaling twelve 5-m by 5-m plots; (B) concrete slab (SLB) plots with 18 blocks in the middle; (C) supporting columns in the middle of the home with crawl space (CRW) plots; and (D) CRW plots. ∆

Soil Science Society of America Journal

5 m (W) space surrounded by stacked concrete blocks. Beside the walls, four vertical columns were installed to aid in supporting the weight of the structure (Fig. 1C). The walls were covered with a wooden deck with doors that provided access to the crawl space for soil sampling (Fig. 1D). The CNT treatment consisted of undisturbed, grassed areas adjacent to the other treatments.

Soil Sampling All twelve plots were sampled (cores of 5-cm diameter) at three depths: 0 to 10 cm, 10 to 20 cm, and 20 to 30 cm before treatment installation. Following installation, samples were collected at 0, 1, 3, 5, 7, 9, and 13 mo. The CNT plots were sampled at three depths (0–10 cm, 11–20 cm, and 21–30 cm) while the soil beneath the two impervious surface treatments were sampled at two depths since the first 10 cm was removed during the construction process. Soil samples were homogenized by sieving through a 2-mm mesh sieve. Bulk density samples were collected next to each sampling point. The bulk density samples were dried at 105°C, and soil bulk density was calculated using the core method (Soil Survey Laboratory Staff, 2004). During the experiment period, no root penetration was observed in CRW and SLB plots. Soil temperature and volumetric water content were measured every 15 min for 15 mo at depths of 10 cm and 20 cm for all treatments using HOBO TidbiT temperature probes (Part UTBI-001, Onset Corp., Pocasset, MA) and EC-5 soil moisture probes (Onset Computer, Bourne, MA), respectively. An automated LabFit AS-3000 pH analyzer was used to measure pH of air-dried soil samples after every sample collection in a 0.01 mol L–1 solution of calcium chloride. This method provides more stable results between seasons compared with the 1:1 mixture of soil/water method (Kissel et al., 2009). The measured values were converted to a 1:1 mixture of soil/water values by adding a conversion factor of 0.6. The Mehlich-1 method ( Jones, 1998) was used to quantify concentrations of extractable P, Ca, and K using inductively coupled plasma spectrograph technique (ICAP61E, Madison, WI). Total C and N were measured using PerkinElmer 2400 Series II CHNS/O analyzer (PerkinElmer, Waltham, MA).

Nitrogen Mineralization Net N mineralization was measured using the in situ bag method (Bottomley et al., 1994). The soil sample was split into two 500-mL polyethylene whirl pack bags (Nasco, Fort Atkinson, WI). Polyethylene bags permit gas exchange while preventing water entering the bags. One bag was immediately reburied and incubated for approximately 28 d, and the other bag was brought back to the laboratory. Post-incubation samples and pre-incubation samples were processed for ammonium and nitrate. Processing consisted of passing the soil through a 2-mm sieve and then extracting 10 g of the 2-mm field moist soil with 100 mL of 2 mol L-1 KCl. The extracted KCl solutions were filtered and then frozen. Extracts were analyzed using standard colorimetric techniques, with the developed color analyzed using a microplate reader (Sims et al., 1995).

Statistical Analysis R version 3.1.2 was used for all statistical analysis in this study (R Core Team, 2014). A linear mixed-effects model (nlme package of R software) with repeated measures of soil properties and blocks as the random effect was utilized to compare soil properties (i.e., soil C, N, microbial biomass C, nitrate, net nitrogen mineralization rates) among treatments. The Tukey’s Honestly Significant Difference (HSD) following analysis of variance (ANOVA) test was used to differentiate the means (Agricola package, a = 0.05).

Results

Soil Volumetric Water and Temperature Soil volumetric water content was impacted by both treatments and soil temperature. In cold periods of the year (November to April) with more precipitation and lower evaporation, soil volumetric water content was higher than in warm periods of the year (May to October) at depth 10 cm, in all treatments (Fig. 2A). However, at depth 20 cm, only the CNT showed a significant difference between warm and cold periods of the year was ob-

Soil Microbial Biomass Soil microbial biomass carbon (Cmic) was measured within 48 h after each sampling. The chloroform-fumigation method was used for measuring Cmic (Vance et al., 1987). Soil samples were homogenized by sieving through a 2-mm mesh sieve and were then divided into two sets of 18.5 g. The first set was fumigated by exposing soil to alcohol-free chloroform for 24 h. Fumigated and non-fumigated samples were extracted with 100 mL of 0.5 mol L–1 K2SO4 for 30 min. The extracts were filtered (Whatman no. 42 filters) and then frozen. After thawing, the samples were analyzed for organic C and total N using a Shimadzu TOC-V and total N combustion analyzer (Shimadzu Scientific Instruments, Columbia, MD). The differences between fumigated and non-fumigated samples represent microbial C and N. www.soils.org/publications/sssaj

Fig. 2. Mean (± standard error, n = 30 per month) monthly soil volumetric water content over a year at (A) 10 cm and (B) 20 cm in control (CNT), concrete slab (SLB), and crawl space (CRW) plots. ∆

Table 1. Mean ± standard error soil temperature (°C) in warm (May to October) and cold periods (November to April) of the year beneath different treatments at depths 10 and 20 cm. Depth of 10 cm Treatment Control (CNT) Concrete slab (SLB) Crawl space (CRW)

May to October

Depth of 20 cm

November to April

May to October

November to April

–––––––––––––––––––––––––––––––––––––––––––––––––– °C –––––––––––––––––––––––––––––––––––––––––––––––––– 25.8 ± 3.4 11.6 ± 5.6 25.6 ± 3.3 11.7 ± 5.3 27.3 ± 3.8 12.0 ± 4.9 27.1 ± 3.7 12.2 ± 4.6 24.6 ± 3.5 11.6 ± 4.1 24.54 ± 3.34 11.8 ± 3.8

served (p = 0.01, F1,10 = 9.85; Fig. 2B). Soil volumetric water content was also impacted by treatments. The CRW treatment had significantly lower water content than other treatments in cold periods of the year (p = 0.02, F2,18 = 4.53), although soil temperature was very similar across all treatments at both depths during that period. In warm periods of the year, SLB treatments had the highest soil volumetric water content among the treatments (p = 0.01, F2,12 = 5.32; Fig. 2A), although soil tempera-

Fig. 3. Mean (± standard error) soil carbon (A) and nitrogen (B) depletion during construction process (month 0, n = 4 per treatment) compared with temporal changes in soil carbon (vs. control plots).

ture beneath the SLB at 10 cm depth was on average 1.45 and 2.66°C warmer than CNT and CRW treatments respectively (Table 1). Higher water content beneath SLB treatments may be due to limited evaporation rates, while higher temperatures beneath SLB are probably due to solar energy absorption of the concrete during the day (Maria et al., 2013).

Soil Carbon and Nitrogen The construction process (Month 0) in which the organicrich, top 10 cm of soil was removed was the primary cause of depletion in carbon content and resulted in 44.00% (±16.16%) and 51.52% (±10.71%) decrease in carbon content beneath SLB and CRW, respectively (Fig. 3A). Afterward, at depth 10 to 20 cm (first layer beneath the concrete slab) soil carbon decreased by 30.4% (±3.4%) from 2.3 (±0.4) kg m-2 at Month 0 to 1.57 (±0.36) kg m-2 at the end of experiment with a rate of 0.04 kg m-2 per month ( ± 0.01, p = 0.001, F1,7 = 9.72; Fig. 4A). Decrease in soil carbon at SLB plots was also observed at depth 20 to 30 cm (second layer beneath the concrete slab) with a rate of 0.02 kg m-2 (±0.005, p < 0.001, F1,6 = 7.93) per month which resulted in 35.2% decrease in soil carbon from Month 0 (1.08 ± 0.3 kg m-2) to the end of experiment (0.70 ± 0.08 kg m-2). Although mean soil carbon content was not significantly different between CNT and SLB plot (Table 2), at the end of experiment (Month 15) soil carbon at SLB plots at depth 10 to 20 cm (1.41 ± 0.25 kg m-2) was 23.8% (±2.7%, p = 0.04, F1,5 = 6.86) lower than that of at CNT plots (2.06 ± 0.39 kg m-2). In contrast to SLB plots, soil carbon at CRW, and CNT did not decrease over time but varied seasonally (Fig. 4B). The soil carbon beneath CRW treatments peaked in November (both 2014 and 2015) at both depths, while the soil carbon in CNT plots reached its maximum in February in the top 10 cm, and did not change over time at depth 10 to 20 cm. Despite the temporal changes, mean soil carbon was not significantly different beneath the SLB (2.31 ± 040 kg m-2) and CRW (1.75 ± 0.19 kg m-2) plots at depth 10 to 20 cm. However, at depth 20 to 30 cm, the soil carbon beneath the CRW (1.14 ± 0.21 kg m-2) was higher than that beneath the SLB (0.93 ± 0.23 kg m-2, p = 0.01, F2,23 = 5.26).

Table 2. Mean ± standard error of soil carbon and nitrogen at control (CNT), concrete slab (SLB) and crawl space (CRW) treatments at depths 0 to 10, 10 to 20, and 20 to 30 cm. Soil carbon

Soil nitrogen

Depth

CNT SLB CRW CNT SLB CRW –––––––––––––––––––––––––––––––––––––––––––––––––––––––––– kg m-2 –––––––––––––––––––––––––––––––––––––––––––––––––––––––––– 0–10 cm 3.94 ± 0.98 N/A† N/A 0.24 ± 0.05 N/A N/A 10–20 cm 1.98 ± 0.41 2.31 ± 0.40 1.75 ± 0.19 0.12 ± 0.05 0.14 ± 0.05 0.11 ± 0.04 20–30 cm 0.99 ± 0.27 0.93 ± 0.23 1.14 ± 0.21 0.06 ± 0.03 0.05 ± 0.03 0.06 ± 0.02 † The top 10 cm was removed during the construction process (beginning of the experiment) of SLB and CRW plots.

∆


Soil carbon content beneath CRW was positively related to soil volumetric water content, and for each one percent increase in soil water content, a 0.23 kg m-2 (±0.008 kg m-2, p = 0.01, F1,18 = 7.06) increase in soil carbon was observed. However, for CNT and SLB plots, no significant relationship between soil water and carbon content was detected. Similar to soil carbon, the construction process resulted in 46.24% (±11.89%), and 58.41% (±3.82%) decreases in soil N in SLB and CRW plots respectively (Fig. 3B). Following that, soil N decreased beneath the SLB over time at both depths. At depth 10 to 20 cm, soil N decreased 4.53 g m-2 (±1.71, p = 0.01, F1,7 = 8.18) for each month of soil sealing. Thus, soil N decreased from 0.14 ± 0.02 kg m-2 at the beginning of the experiment to 0.06 ± 0.01 kg m-2 at Month 15 after soil sealing. Similarly, Fig. 4. Temporal changes of mean (± standard error) soil carbon (n = 4 per month) and nitrogen (n = 4 at depth 20 to 30 cm soil N de- per month) beneath concrete slab (SLB; A, C), and crawl space (CRW; B, D) at 10 to 20 cm. The number creased 3.01 g m-2 (±1.09, p = 0.01, in the parenthesis in the x-axis refers to the month following the start of the project. F1,6 = 7.66) for each month of soil sealing (Fig. 4C). In contrast to SLB plots, soil N at CRW and consequently (p = 0.12), while Cmic had seasonal variation in CNT plots fluctuated seasonally but did not show a temporal CRW plots at depth 10 to 20 cm (Fig. 5A). Microbial biomass pattern (Fig. 4D). carbon was significantly related to soil volumetric water content The average soil C to N ratio at a depth of 10 to 20 cm foland increased 34.99 mg g-1 (±17.39 mg g-1, p = 0.04, F1,18 = 8.86), lowed the order of CRW (20.9 ± 2.9) > CNT (19.9 ± 1.9) > and for each one percent increase in soil water at depth 10 to 20 cm. SLB (17.2 ± 2.8), however, the difference between treatments was The Cmic decreased with depth for all treatments (p < 0.001). not statistically significant (p = 0.4, F2,24 = 0.27). A similar trend However, at depth 20 to 30 cm, Cmic at CRW plots was signifiwas observed at depth 0 to 10 cm.

Soil Microbial Biomass Soil microbial biomass (Cmic) at 10 to 20 cm was not statistically different among the treatments and followed the order of CRW (348.47 ± 48.16), SLB (260.86 ± 66.74), and CNT (230.01 ± 65.93; Fig. 5A). The Cmic at SLB plots declined with a month lag after construction which may be due to extra moisture provided by wet concrete at the beginning of the construction process (Fig. 5A). After a month, the Cmic significantly decreased (p > CRW > CNT treatments (p < 0.001, F2,24 = 5.38). Initial soil nitrate contents decreased with depth in CNT and CRW plots (p CRW (2.33 ± 10.12 mg N m-2 d-1; Fig. 6A). The warm period of the year showed significantly higher net nitrogen mineralization rates than the cooler periods (p < 0.001). The highest increase in mineralization rates in warm periods of the year was observed beneath SLB such that in May and July, the nitrogen mineralization rates beneath the SLB were higher than that in CNT plots. Net ammonium production beneath the SLB and CRW was near zero or negative. Thus, the

Mean soil pH increased underneath SLB and CRW plots and followed the order of SLB (6.13 ± 0.40) » CRW (6.05 ± 0.35) » CNT (5.38 ± 0.11) at depth 10 to 20 cm. Soil pH increased by 0.088 (±0.016, p = 0.001, F1,7 = 9.3) per month beneath SLB and 0.028 (±0.008, p < 0.001, F1,7 = 9.6) per month beneath the CRW during the experiment time frame but did not show a temporal pattern at CNT plots. As a result soil pH at SLB plots increased from 5.14 ± 0.07 at the beginning of the experiment to 5.90 ± 0.17 at the end of the experiment. Soil pH decreased with depth in all treatments and at depth 20 to 30 cm had a pattern similar to that of 10 to 20 cm. Soil potassium (K) concentrations at depth 10 to 20 cm were significantly higher in SLB plots compared with the CNT and CRW treatments (p < 0.001, F 2,24 = 10.31; Fig. 7). Temporally, for each month, soil K increased 23.10 mg kg-1 (±4.40, p < 0.001, F1,7 = 9.02) beneath SLB while no significant change was observed over time in the other treatments. Potassium accumulation beneath SLB was only observed at depth 10 to 20 cm which is in contact with impervious surfaces. Soil calcium (Ca) concentrations in SLB plots at depth 10 to 20 cm were significantly higher than that in CRW and CNT plots (p = 0.003, F2,24 = 8.17). There was no significant difference among treatments in Ca concentrations at depth 20 to 30 cm. While the soil phosphorus (P) content was not statisti-

Fig. 7. Mean (± standard error) soil Ca, K, and P concentration beneath control (CNT), concrete slab (SLB), and crawl space (CRW) treatments. Means are significantly different with different letters based on Tukey’s honest significant difference (HSD) tests at a = 0.05. ∆


cally different among the treatments, soil P concentrations increased beneath the CRW by 0.12 mg kg-1 (±0.06, p = 0.05, F1,7 = 6.8) but there was no significant relationship between time and soil P in other treatments.

Discussion The decrease in soil carbon and nitrogen content beneath both impervious surfaces was primarily due to top soil removal and initial disturbance as opposed to post-construction processes. Removal of the top 10 cm of soil during the construction process (Month 0) resulted in 44.00% (±16.16%) and 51.52% (±10.71%) decreases in carbon content beneath SLB and CRW, respectively. The results of this study suggest that although a significant decrease in carbon and nitrogen has been observed in field studies, it cannot simply be translated to carbon loss to the atmosphere (Wei et al., 2014b, 2014a; Yan et al., 2015; Majidzadeh et al., 2017). The removed top soil during initial disturbances may have been reused, and if not, it would not lose all the carbon immediately. For instance, Raciti et al. (2012) reported a 66% decrease in carbon beneath pavements compared with nearby soils in the top 15 cm, and from that concluded that 3.2 and 0.4 Pg carbon may have been lost to atmosphere from the soil beneath impervious surfaces of the world and United States, respectively. However, by taking into account the top soil removal, carbon loss from soil beneath impervious surfaces would not exceed 1.92 Pg globally and 0.24 Pg in the United States, considering that at least 40% of that loss is due to the top soil removal. Similarly, Majidzadeh et al. (2017) studied soil carbon beneath homes that ranged from 11 to 114 yr in age built with crawl spaces in Alabama and Georgia and reported a 61.86% (±4.42%) decrease in carbon beneath homes compared with adjoining urban lawns. Another study in Nanjing, China reported a 68% carbon depletion in the top 20 cm beneath paved residential squares (Wei et al., 2013). The highest carbon loss beneath impervious surfaces was reported by Yan et al. (2015), who noted a 75% decrease in carbon content in the top 10 cm. However, the latter study was in a dry region which reduced the input of organic matter to a major extent. Carbon contents beneath CRW and SLB were not significantly different at depth 10 to 20 cm although CRW treatments exhibited higher carbon at depth 20 to 30 cm, suggesting that the soil beneath the CRW may not be as separated from inputs as that under SLB. This was further supported by higher volumetric water content beneath the CRW at depth 20 to 30 cm and a significant relation between soil volumetric water and carbon content underneath CRW (p = 0.01, F1,18 = 7.06). Soil Cmic in SLB plots decreased over time as expected considering the lack of bioavailable C and N (Wardle, 1992). Moreover, gas and water exchange was limited by soil sealing which likely reduced microbial activities. However, while Cmic declined over time at SLB plots, it fluctuated in the CRW treatment. This was probably due to limited inputs of organic carbon which would have enabled some microbial activity there. This was further confirmed by a significant relationship between soil www.soils.org/publications/sssaj

volumetric water content and Cmic which is in accordance with Piotrowska-Długosz and Charzyński (2015) who showed that soil moisture and microbial activity could be affected by the degree of sealing. A higher Cmic to C ratio was observed beneath the CRW compared with the other treatments at depth 10 to 20 cm. This ratio has a positive relationship with the proportion of metabolized carbon (Scharenbroch et al., 2005), thus the increase in the Cmic to C ratio suggests C loss over time (Beyer et al., 1995). Soil carbon beneath SLB treatment showed a significant relation to Cmic, while soil carbon content beneath CRW had a significant relationship with soil volumetric water content. Net nitrogen mineralization rates were affected by both temperature and treatments. Higher net nitrogen mineralization during the warm periods of the year, however, SLB plots exhibited such large increases in May and July that mineralization rates beneath SLB were higher than those in CNT plots. This is probably due to higher water content beneath SLB compared with CNT plots during warm periods of the year. Net ammonium production was close to zero beneath impervious surfaces since organic matter inputs were almost eliminated. However, initial soil ammonium content was sufficient to support nitrate production. This pattern will likely subside through time since initial organic matter would not be replenished. Wei et al. (2014b) also reported that total soil mineral nitrogen (NO3- and NH4+) concentrations were lower beneath pavements compared with open soils. Soil pH increased over time at depth 10 to 20 cm in both SLB and CRW treatments. P accumulation beneath the CRW treatments may be a result of the increase in soil pH beneath the impervious surfaces. Moreover, the P accumulation may be due to the absence of P uptake by plants (Soldat and Petrovic, 2008) and/or reduced erosion losses of P via runoff (Wei et al., 2014b). Results of this study indicate that soil carbon and nitrogen dynamics beneath structures built on crawl spaces vs. concrete slabs) are not the same and partially permeable impervious surfaces such as CRW may support inputs of carbon and organic matter even in short periods after soil sealing. Besides the type of structure, soil biogeochemical properties beneath impervious may also be impacted by other parameters that are not monitored in this study such as degree of topsoil removal, the age of structure, surrounding vegetation and trees, and climate. Results from this study suggest that carbon loss from soil beneath impervious surfaces would not exceed l.92 Pg globally and 0.24 Pg in the United States.

Acknowledgments

This work supported by the Center for Environmental Studies at the Urban-Rural Interface (CESURI), School of Forestry and Wildlife Sciences, Auburn University. We gratefully acknowledge Dr. Todd Steury for providing feedback on an early draft.

References

Beyer, L., H.-P. Blume, and D.-C. Elsner. 1995. Soil organic matter composition and microbial activity in urban soils. Sci. Total Environ. 168:267–278. doi:10.1016/0048-9697(95)04704-5 Bottomley, P.S., J.S. Angle, R.W. Weaver, S.C. Hart, J.M. Stark, E.A. Davidson,

∆

and M.K. Firestone. 1994. Nitrogen mineralization, immobilization, and nitrification. In: R.W. Weaver, et al., editors, Methods of soil analysis, part 2. SSSA Book Series. SSSA, Madison, WI. p. 985–1018. Charzyński, P., P. Hulisz, A. Piotrowska-Długosz, D. Kamiński, and A. Plak. 2017. Sealing effects on properties of urban soils. In: R. Lal, and B.A. Stewart, editors, Urban soils. Advances in soil science. Taylor & Francis Group, Boca Raton, FL. p. 155–174. Elvidge, C.D., C. Milesi, J.B. Dietz, B.T. Tuttle, P.C. Sutton, R. Nemani, and J.E. Vogelmann. 2004. US constructed area approaches the size of Ohio. Eos Trans. AGU 85:233–233. doi:10.1029/2004EO240001 IUSS Working Group WRB. 2015. World reference base for soil resources 2014: International soil classification system for naming soils and creating legends for soil maps. World Soil Resources Reports No. 106. FAO, Rome. Jones, J.B. 1998. Soil test methods: Past, present, and future use of soil extractants. Commun. Soil Sci. Plant Anal. 29:1543–1552. doi:10.1080/00103629809370048 Kissel, D.E., L. Sonon, P.F. Vendrell, and R.A. Isaac. 2009. Salt concentration and measurement of soil pH. Commun. Soil Sci. Plant Anal. 40:179–187. doi:10.1080/00103620802625377 Lal, R. 2004. Soil carbon sequestration impacts on global climate change and food security. Science 304:1623–1627. doi:10.1126/science.1097396 Lehmann, A., and K. Stahr. 2007. Nature and significance of anthropogenic urban soils. J. Soils Sediments 7:247–260. Lorenz, K., and R. Lal. 2009. Biogeochemical C and N cycles in urban soils. Environ. Int. 35:1–8. doi:10.1016/j.envint.2008.05.006 Majidzadeh, H., B.G. Lockaby, and R. Governo. 2017. Effect of home construction on soil carbon storage: A chronosequence case study. Environ. Pollut. 226:317–323. doi:10.1016/j.envpol.2017.04.005 Maria, V.D., M. Rahman, P. Collins, G. Dondi, and C. Sangiorgi. 2013. Urban heat island effect: Thermal response from different types of exposed paved surfaces. Int. J. Pavement Res. Technol. 6:414–422. Mendyk, Ł., and P. Charzyński. 2016. Soil sealing degree as factor influencing urban soil contamination with polycyclic aromatic hydrocarbons (PAHs). Soil Sci. Annual 67:17–23. doi:10.1515/ssa-2016-0003 Montague, T., and R. Kjelgren. 2004. Energy balance of six common landscape surfaces and the influence of surface properties on gas exchange of four containerized tree species. Sci. Hortic. (Amsterdam) 100:229–249. doi:10.1016/j.scienta.2003.08.010 Morel, J.L., C. Chenu, and K. Lorenz. 2015. Ecosystem services provided by soils of urban, industrial, traffic, mining, and military areas (SUITMAs). J. Soils Sediments 15:1659–1666. Piotrowska-Długosz, A., and P. Charzyński. 2015. The impact of the soil sealing degree on microbial biomass, enzymatic activity, and physicochemical properties in the Ekranic Technosols of Toruń (Poland). J. Soils Sediments 15:47–59. doi:10.1007/s11368-014-0963-8 Pouyat, R.V., I.D. Yesilonis, and D.J. Nowak. 2006. Carbon storage by urban soils in the United States. J. Environ. Qual. 35:1566–1575. doi:10.2134/

∆

jeq2005.0215 R Core Team. R software, version 3.1.2. R Foundation for Statistical Computing, Vienna, Austria. Raciti, S.M., L.R. Hutyra, and A.C. Finzi. 2012. Depleted soil carbon and nitrogen pools beneath impervious surfaces. Environ. Pollut. 164:248– 251. doi:10.1016/j.envpol.2012.01.046 Scalenghe, R., and F.A. Marsan. 2009. The anthropogenic sealing of soils in urban areas. Landsc. Urban Plan. 90:1–10. doi:10.1016/j. landurbplan.2008.10.011 Scharenbroch, B.C., J.E. Lloyd, and J.L. Johnson-Maynard. 2005. Distinguishing urban soils with physical, chemical, and biological properties. Pedobiologia 49:283–296. doi:10.1016/j.pedobi.2004.12.002 Sims, G.K., T.R. Ellsworth, and R.L. Mulvaney. 1995. Microscale determination of inorganic nitrogen in water and soil extracts. Commun. Soil Sci. Plant Anal. 26:303–316. doi:10.1080/00103629509369298 Soil Survey Laboratory Staff. 2004. Survey laboratory methods manual. USDA, Lincoln, NE. Soldat, D.J., and A.M. Petrovic. 2008. The fate and transport of phosphorus in turfgrass ecosystems. Crop Sci. 48:2051. doi:10.2135/cropsci2008.03.0134 US Census Bureau. 2013. American housing survey. US Census Bureau, Washington, DC. http://www.census.gov/programs-surveys/ahs.html (Verified 19 June 2016). Vance, E.D., P.C. Brookes, and D.S. Jenkinson. 1987. Microbial biomass measurements in forest soils: The use of the chloroform fumigationincubation method in strongly acid soils. Soil Biol. Biochem. 19:697–702. doi:10.1016/0038-0717(87)90051-4 Wardle, D.A. 1992. A comparative assessment of factors which influence microbial biomass carbon and nitrogen levels in soil. Biol. Rev. Camb. Philos. Soc. 67:321–358. doi:10.1111/j.1469-185X.1992.tb00728.x Wei, Z., S. Wu, and C. Lin. 2013. Installation of impervious surface in urban areas affects microbial biomass, activity (potential C mineralization), and functional diversity of the fine earth. Soil Res. 51:59. doi:10.1071/ SR12089 Wei, Z., S. Wu, X. Yan, and S. Zhou. 2014a. Density and stability of soil organic carbon beneath impervious surfaces in urban areas. PLoS One 9:E109380. doi:10.1371/journal.pone.0109380 Wei, Z., S.-H. Wu, S.-L. Zhou, J.-T. Li, and Q.-G. Zhao. 2014b. Soil organic carbon transformation and related properties in urban soil under impervious surfaces. Pedosphere 24:56–64. doi:10.1016/S1002-0160(13)60080-6 Yan, Y., W. Kuang, C. Zhang, and C. Chen. 2015. Impacts of impervious surface expansion on soil organic carbon– A spatially explicit study. Sci. Rep. 5:17905. doi:10.1038/srep17905 Zhu, W., P. Hulisz, B.A. Egitto, I.D. Yesilonis, and R.V. Pouyat. 2017. Soil carbon and nitrogen cycling and ecosystem service in cities. In: R. Lal, and B.A. Stewart, editors, Urban soils. Advances in soil science. Taylor & Francis Group, Boca Raton, FL. p. 121–136.
