Development of Sandy Soil Properties Over Time within Subtropical Residential Landscapes

doi:10.21203/rs.3.rs-2869877/v1

Download PDF

Research Article

Development of Sandy Soil Properties Over Time within Subtropical Residential Landscapes

https://doi.org/10.21203/rs.3.rs-2869877/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Urbanization is homogenizing residential developments and soils across a wide climate gradient of the United States. Land development practices commonly install landscapes on low quality compacted fill material. While previous research shows that soil quality beneath urban lawns improves over time, studies used golf courses to represent lawns or sampled across variable development and management styles. Limited data on sandy soil quality within residential landscapes in subtropical climate warranted investigation. To evaluate the effect of age on residential soil characteristics, two residential developments from Central and Southwest Florida were selected based on known lot ages up to 31 years. Soil samples (0 to 30 cm) were collected from landscape beds and lawns, along with lawn of common areas, to evaluate bulk density and organic matter over time. Though landscape beds typically require less management, soil characteristics were similar between lawns and landscape beds, suggesting similar processes. Bulk densities decreased by 22–34% and 9–20% within the upper (0–15 cm) and lower (15–30 cm) soil depths, respectively. Upper layer soil carbon accumulation rates ranged from 0.060–0.156 kg C m^− 2 yr^− 1, comparable to previous studies but did not exceed 0.021 kg C m^− 2 yr^− 1 or were not significant for greater depths. The results of this study are representative of past chronosequence studies conducted in cooler climates with fine soil texture. Sandy soil texture and warm climate do not govern carbon sequestration rates, but instead vegetation type and landscape management practices overcome these conditions.

Urban soil

residential landscapes

chronosequence

turfgrass

carbon accumulation

soil quality

Urbanization is homogenizing areas across the moisture gradient of the United States (Groffman et al. 2014) and residential soils are specifically homogenizing (Ryan et al. 2022). Land development has an impact on ecosystems, and in particular, soil quality. In residential developments, soil quality is often degraded by removal of predevelopment topsoil, regrading with low quality fill dirt, and compaction by heavy machinery traffic during construction (Kozlowski 1999; Chen et al. 2014). Compaction and the low organic matter content in urban soils reduces plant-available water storage, reduces infiltration, and increases surface runoff, compared to soils which are undisturbed (Jim 1998; Pitt et al. 1999; Cogger 2005; Gregory et al. 2006). Additionally, sandy soils common to many regions are especially susceptible to reduced infiltration due to compaction (Pitt et al. 1999; Gregory et al. 2006).

Residential landscapes are commonly composed of some combination of turfgrass and landscape beds. Milesi et al. (2005) estimated turfgrass covers 1.9% of the United States land areas, as it is effective for erosion control and costs less than landscape bed plantings per unit area installed. Florida residents use over half of domestic freshwater consumption outside the home for landscape irrigation (Marella 2020). Despite receiving a statewide average of 137 cm of precipitation a year (Florida State University 2023), Florida uses the fourth most water (11,500 Mgal d^-1) of any state in the United States, and the most of any state east of the Mississippi River (Dieter et al. 2018). Florida’s population and number of visitors have increased over time to 21.5 million residents in 2020 (US Census Bureau 2023) and to 131 million out-of-state visitors in 2019 (Visit Florida 2023), respectively. As a result, in 2015, Florida’s domestic freshwater use (2,215 Mgal d^-1) surpassed agricultural use (2,089 Mgal d^-1; Marella 2020) for the first time in recorded history.

Soil quality may be defined as "the capacity of a specific kind of soil to function, within natural or managed ecosystem boundaries, to sustain plant and animal productivity, maintain or enhance water and air quality, and support human health and habitation.” (Karlen et al. 1997). Past studies have shown that soil quality tends to improve over 20–55 years in urban, turfgrass-based systems as organic matter accumulates (Qian and Follett 2002; Raciti et al. 2011; Selhorst and Lal 2011; Huyler et al. 2014). Accumulation of organic matter and changing soil structure can affect soil quality and resources needed to support landscape vegetation. These studies took place across a variety of climates and locations within the United States, including Colorado and Wyoming (Qian and Follett 2002; putting greens), Maryland (Raciti et al. 2011; lawns), Ohio (Selhorst and Lal 2011; golf courses), and Alabama (Huyler et al. 2014; untreed lawns). However, these studies were either from golf courses or from lawns across municipal areas with varying land development and landscape management practices. No previous studies have investigated the evolution of soil quality over time under both lawns and landscape beds within the constraints of individual residential developments, to minimize confounding factors of soils, development practices, and landscape management practices.

Further, soil textures were variable among previous studies, with most sites characterized by loamy soils and only one site (Qian and Follet 2002; putting greens) exhibiting sandy soils. Sandy soils are common throughout the state of Florida (Fig. 1), and the Candler series (hyperthermic, uncoated Lamellic Quartzipsamments) is the most widespread soil in the state (Soil Survey Staff 2021). The combination of soil and climate results in limited organic matter content of natural soils. This contrasts with the previous studies from other regions of the US that showed organic matter increased with time.

Given Florida’s expected population growth, and the growth of urban areas worldwide, it is important to understand the impact of development on urban soils and their future evolution.

A scarcity of evidence for improvement to soil quality over time in urban developments with sandy soils or a sub-tropical climate warranted study of this phenomenon within the state of Florida where accumulation of soil organic matter in undisturbed soils is limited. The overall goal of this study was to assess soil quality (0 to 30 cm) change over time since initial disturbance in two residential developments under lawns and landscape beds, located in Central and Southwest Florida. in regions. Specifically, the objectives were to: (1) identify trends in soil properties (bulk density, organic matter content, and organic carbon accumulation) over time for three land cover types (landscape beds, lawns, and common areas); (2) compare and contrast analyzed relationships based on sample collection sites (Central or Southwest FL), location (landscape beds, lawns, common areas), and sample depth to determine similarities and differences.

Site Descriptions

Lots were sampled from two urban residential developments in Florida (Fig. 2): On Top of the World (OTOW) in Central Florida (latitude 29.1°N longitude 82.3°W), and Lakewood Ranch (LR) in Southwest Florida (latitude 27.4°N longitude 82.4°W). Mean annual rainfall for OTOW is 1,318 mm and mean annual temperatures is 21°C; for LR, mean annual rainfall is 1,377 mm, and mean annual temperature is 23°C (NOAA 2021). Prior to development most areas within OTOW were primarily mapped as the Candler soil series (hyperthermic, uncoated Lamellic Quartzipsamments; Soil Survey Staff 2021), the most common soil series in Florida. Prior to development in LR areas were mapped as either the Myakka series (sandy, siliceous, hyperthermic Aeric Alaquods) or the Eau Gallie series (sandy, siliceous, hyperthermic Alfic Alaquods; Soil Survey Staff, 2021). However, published soil surveys do not reflect potential additions of fill or other landscape modifications that may have occurred during home construction (Gregory et al., 2006) as soil surveys were published for Marion County (OTOW) and Manatee County (LR) in 1978 and 1980, respectively; before either development began construction.

Residential lots in LR (LR-R; n = 50) were sampled in 2020 and 2017 for OTOW (n = 45), while the LR common areas (LR-CA; n = 16) were sampled in 2015 (Table 1). Lot age and common area age were calculated based on the year the property was developed and the year the property was sampled. The development date of each residential lot was verified using the Marion and Manatee County Property Appraiser’s databases, respectively, for OTOW and LR (Manatee Country Property Appraiser 2021; Marion County Property Appraiser 2022). Lot ages ranged from 0 to 30 years since development (1986–2016) for LR-R locations, 0 to 20 years (1995–2015) for LR-CA, and 0 to 22 years for OTOW (1998–2020).

Table 1

The number of lots sampled is within each lot age range for two residential developments in Florida. The sampling year and the age of the oldest lot at the time of sampling were used to calculate the lot age.
Development			Sampling Year	Oldest Lot	Number of lots sampled by age range
Name	Lot Type	Acronym	Sampling Year	Oldest Lot	0–5	6–10	11–20	≥ 21	Total
On Top of the World	Residential	OTOW	2017	1986	19	6	10	10	45
Lakewood Ranch	Residential	LR-R	2020	1998	10	10	22	8	50
Lakewood Ranch	Common Area	LR-CA	2015	1995	9	0	7	0	16

For OTOW and LR-R, between six and 19 lots were selected within five age group categories: 0–5 yrs., 6–10 yrs., 11–15 yrs., 16–20 yrs., ≥ 21 yrs. OTOW had a total of 45 sites while LR-R had 50 sites. For LR-CA, three to four locations were sampled from within areas 0, 1, 5, 13, and 20 years after development, for a total of 111 sites.

Soil Sampling

For LR-CA, a soil core sampler was used to collect samples from 0–6 cm and 6–30 cm depths. For the 6–30 cm depth, the sampling started at the 15 or 16 cm depth to capture the midpoint of this subsurface layer. A minimum of two subsamples per depth were taken within each common area.

For OTOW and LR-R survey flags were placed in random locations throughout the lawn and landscape beds to mark sampling locations (Fig. 3-A). In each lot, subsamples were collected from five locations in the front lawn and five locations in the front landscape beds, excluding any areas between the sidewalk and the street. Landscape beds containing utility boxes, mature trees, gravel, or landscaping fabric were avoided. Mulch in landscape beds was pulled aside to allow for penetration of the soil sampler. For OTOW only, soil samples were collected from new lots (Age = 0 yrs.), just before installation of turfgrass or landscape beds. Nine total sampling locations were randomly selected within the new lots, and the mean of the nine subsamples was used to represent one lot.

Using a soil core sampler to determine bulk density at multiple depths is destructive (Al-Shammary et al. 2018), and was therefore impractical for collecting samples in the LR-R and OTOW settings. Instead, soil samples were collected using narrow, tubular soil samplers, or punch tubes. For LR-R, a 56 cm punch tube with an internal cutting diameter of 1.9 cm was used to collect samples. In OTOW, a 121 cm punch tube with an internal cutting diameter of 1.7 cm was used for soil sample collection (18 from pre landscaping and 144 from aged lots). In both OTOW and LR residential areas, the punch tube was pushed into the ground to a depth of 30 cm or until refusal, and the depth to refusal was recorded (Fig. 3-B).

Occasionally, high levels of compaction, tree roots, rocks, or other obstructions were prohibitive to collecting a sample deeper than 15 cm. If a sample at a depth of at least 15 cm could not be obtained, a new sampling point within 1–2 m of that was selected for resampling. Total sampling depth was recorded to account for samples ranging in total depth from 15 to 30 cm.

Each soil sample was divided into an upper 15 cm section and a lower section of up to 15 cm. Any surface vegetation or turf thatch were cut from the lawn-collected samples before dividing into upper and lower depths (Fig. 3-C). The five subsamples taken at each location (lawn or landscape bed) and depth (0–15 cm or 15–30 cm) were composited by sample location and depth (e.g., lawn & 0–15 cm depth) for each lot, which yielded four composite samples per lot.

Soil Sample Analyses

Bulk Density

All soil samples were oven dried at 105°C for 48 hours or until weights no longer declined. Dry bulk density was calculated using the sieved (#10; < 2 mm) oven-dry mass and cylindrical volume of either (1) soil plugs (cutting diameter and length) collected with the punch tube or (2) soil core collected with soil core sampler. While Gravels and other rock fragments in the soil were infrequent, > 2 mm-sized particles were removed from the bulk density determination.

Soil Texture

Soil particle size distributions for soil samples from OTOW and LR-R sites were determined by laser diffraction analysis, a volumetric determination of particle size, using a Beckman Coulter LS-13320 multi‐wave particle size analyzer at the University of Florida Environmental Pedology and Land Use Laboratory. Subsamples from composite samples were analyzed without dispersion or pre-treatment (Pachon et al, 2016). Particle size distributions are summarized here as the volumetric concentrations of sand- (2.0–0.05 mm), silt- (0.05–0.002 mm), and clay-sized (< 0.002 mm) particles. For the LR common areas, subsamples were analyzed for sand, silt, and clay-sized particles using the hydrometer method (Gee and Bauder, 1986). The textural class for each sample was determined using the relative amounts of each particle size (USDA, 1987).

Soil organic matter

Soil organic matter (SOM) concentration was determined by loss on ignition (LOI) with an ignition temperature of 550°C and a total ignition time of 4 hours. Soil organic carbon (SOC) was estimated from SOM as 0.50*SOM (Pribyl 2010; Weil and Brady 2017). Carbon density was then calculated by multiplying SOC by bulk density and depth of the soil sample. A carbon accumulation rate was estimated by the difference in mean carbon density between 0- and 22-year-old lots for LR-R and 0- and 30-year-old lots for OTOW.

Data Analysis

Linear regression models were developed using R (R Core Team 2021) to test for associations between lot age and soil organic matter, and bulk density. Separate linear models were generated for soil samples collected from lawns and landscape beds, as well as for the upper and lower portions of the collected soil profile. Assumptions of normality were assessed by visual inspection of a normal Q-Q plot and histogram of residual values, and by the Shapiro-Wilk test. Homoscedasticity was evaluated visually by plotting model residuals versus predicted values and by the Breusch–Pagan test.

In several cases, the statistical significance of the linear regression model was evaluated on transformations of the dependent variable to satisfy linear model assumptions. The Box-Cox method (Box and Cox 1964) was used to select the most appropriate power transformation (e.g. log, square root, cubic, etc.). Simple linear regression assumptions could not be satisfied by transformation in the analysis of soil texture, so the Thiel-Sen estimator method, a nonparametric assessment of linear trends, was used instead of a simple linear regression (Kumar Sen 1968). Base R (R Core Team, 2021) and packages ‘car’ (Fox & Weisburg 2019), ‘ggplot2’ (Wickham 2016), ‘ggpubr’ (Kassambara 2020), ‘lmtest’ (Zeileis & Hothorn 2002), ‘MASS’ (Venables & Ripley 2002), and ‘mblm’ (Komsta 2019) were used for data organization, analysis, modeling, and visualization.

Soil Particle Size

All soil samples from LR-R lots were dominated by sand-sized particles. While texture is strictly based on particle mass, rather than volume, textures were estimated to put soils into context. For most soil size fractions (sand, silt, clay), results of the Thiel-Sen estimator method indicated a significant association (P < 0.05) between lot age (years) and particle size. However, any differences in soil texture observed over the range of lot ages represented in this study (0–22 years) were not large enough to change textural classifications. All soils collected from LR-CA had a sand texture with approximately 1% clay.

Soil samples from OTOW and LR typify the coarse-textured nature of FL soils (Table 2). Volumetric concentration of sand-sized particles averaged more than 95% from 0–15 cm in both lawns and landscape beds at both residential developments, and always exceeding 84%. From 15–30 cm, the concentration of sand-sized particles averaged 94% in lawns and landscape beds of both residential developments, and never less than 78%. These volumetric concentration estimates correspond to the USDA texture classes of sand and loamy sand as defined gravimetrically on the soil texture triangle.

Table 2

Volumetric concentration (%) of sand-, silt-, and clay-sized particles of residential lots sampled in On Top of the World and Lakewood Ranch. Values are shown as means ± standard deviations with the range listed below as *minimum – maximum.*
	OTOW			LR-R
	Sand	Silt	Clay	Sand	Silt	Clay
	0–15 cm
Lawn	96.1 ± 1.6	3.2 ± 1.5	0.8 ± 0.1	95.6 ± 1.6	3.5 ± 1.4	0.9 ± 0.2
	90.5–98.0	1.3–11.4	0.7–1.8	90.8–98.0	1.3–7.9	0.6–1.4
Landscape Bed	95.7 ± 2.1	3.4 ± 2.0	0.9 ± 0.2	95.6 ± 1.7	3.5 ± 1.6	0.9 ± 0.2
Landscape Bed	88.4–97.7	1.4–10.2	0.6–1.4	87.7–97.9	1.5–10.5	0.5–1.8
	15–30 cm
Lawn	93.7 ± 4.5	5.3 ± 4.2	1.0 ± 0.3	94.1 ± 2.4	4.8 ± 2.1	1.0 ± 0.3
	78.3–97.5	1.5–20.0	0.7–1.7	84.9–97.2	2.1–12.9	0.7–2.1
Landscape Bed	94.5 ± 3.1	4.4 ± 2.8	1.1 ± 0.3	94.2 ± 2.1	4.8 ± 1.9	1.0 ± 0.2
Landscape Bed	87.0–97.8	1.3–11.4	0.7–1.8	88.8–97.2	2.0–9.6	0.7–1.6

Bulk Density

OTOW

For OTOW, increased lot age was associated with a decreased soil bulk density for all sampling depths and locations (p < 0.05), and changes in bulk density with lot age had a higher adjusted R² compared to trends observed in LR-R (Fig. 4). For lawn-collected soil samples (Fig. 4a), lot age significantly predicted soil bulk density at a depth of 0–15 cm (inverse cube root transformation; F(1, 43) = 98, P = 1.1*10^–12, R²_adj = 0.69) and at a depth of 15–30 cm (F(1, 42) = 54, P = 4.7*10^− 9, R²_adj = 0.55). For soil samples collected from landscape beds (Fig. 4c), lot age significantly predicted soil bulk density at a depth of 0–15 cm (F(1, 43) = 49, P = 1.1*10^− 8, R²_adj = 0.52) and at a depth of 15–30 cm (F(1, 42) = 37, P = 3.4*10^− 7, R²_adj = 0.45). Additionally, the regression models suggest that over 30 years, mean bulk density at a depth of 0–15 cm decreased by 0.41 and 0.39 g cm^-3 (34% and 32% change) in lawns and landscape beds, respectively. Bulk density decreased to a lesser extent at a depth of 15–30 cm, with a mean decrease of 0.38 and 0.39 g cm^-3 (19 and 20% decrease) for lawns and landscape beds, respectively.

LR-R

Increased lot age was associated with a decrease in soil bulk density for all sampling depths and locations (P < 0.05), with weaker associations for samples collected at a depth of 15–30 cm. For lawn-collected soil samples (Fig. 4b), lot age significantly predicted soil bulk density at a depth of 0–15 cm (F (1, 48) = 41.48, P < 0.0001, R²_adj = 0.45) and at a depth of 15–30 cm (cubic transformation; F (1, 48) = 16.95, P < 0.0001, R²_adj = 0.24). For soil samples collected from landscaping beds (Fig. 4d), lot age significantly predicted soil bulk density at a depth of 0–15 cm (F(1, 48) = 22.64, P < 0.0001, R²_adj = 0.31) and at a depth of 15–30 cm (cubic transformation; F (1, 48) = 5.314, P = 0.03, R²_adj = 0.08). Additionally, the regression models suggest that over 22 years, mean bulk density at a depth of 0–15 cm decreased by 0.54 and 0.35 g cm^-3 (31 and 24% decrease) in lawns and landscape beds, respectively. Bulk density decreased to a lesser extent at a depth of 15–30 cm, with a mean decrease of 0.33 and 0.19 g cm^-3 (15 and 9% decrease) for lawns and landscape beds, respectively.

LR-CA

In LR common areas, decreased bulk density was associated with an increase in development age (Fig. 5a) at a depth of 0–6 cm (F (1, 35) = 53.78, P = 1.4*10^− 8, R²_adj = 0.59) and a depth of 6–30 cm (reciprocal square transformation; F (1, 25) = 25.71, P = 3.1*10^− 5, R²_adj = 0.49). Mean bulk density decreased from 1.67 to 1.30 g cm^-3 (22% decrease) between 0- and 20-year-old common areas.

Soil Organic Matter

OTOW

For lawn-collected soil samples (Fig. 6a) at a depth of 0–15 cm, lot age significantly predicted a reciprocal cube root transformation of SOM (F(1, 43) = 88, P = 6.3*10^–12, R²_adj = 0.66), but the relationship between lot age and SOM was not significant at a depth of 15–30 cm (P = 0.995). For samples collected from landscape beds (Fig. 6c), lot age significantly predicted a reciprocal square root transformation of SOM at a depth 0–15 cm (F(1, 43) = 33, P = 7.5*10^− 7, R²_adj = 0.42), but similarly to lawn-collected samples, the relationship between lot age and SOM was not significant at a depth of 15–30 cm (P = 0.524). These results also suggest that in the upper 15 cm of the soil profile, SOM increased by approximately 5.7% and 2.5% (384% and 194% increase) in lawns and beds, respectively, between 0 and 30 years after lot development.

LR-R

Overall, increased lot age was associated with a significant (α = 0.05) increase in SOM for all sampling depths and locations, with weak associations for samples collected at a depth of 15–30 cm. For lawn-collected soil samples at a depth of 0–15 cm (Fig. 6b), lot age significantly predicted SOM (log transformation; F (1, 48) = 57.08, P < 0.0001, R²_adj = 0.53), and an inverse cube root transformation of SOM at a depth of 15–30 cm (F(1, 48) = 7.602, p = 0.008, R²_adj = 0.12). For soil samples collected from landscape beds (Fig. 6d), lot age significantly predicted a log transformation of SOM at a depth 0–15 cm (F (1, 48) = 50.8, p < 0.0001, R²_adj = 0.50) and 15–30 cm (F (1, 48) = 4.27, p = 0.04, R²_adj = 0.06). These results also suggest that in the upper 15 cm of the soil profile, SOM increased by approximately 2.7% and 3.2% (163% and 158% increase) in lawns and beds, respectively, between 0 and 22 years after lot development. In contrast, an increase in SOM of only 0.53 and 0.44% (53% and 36% change) was respectively observed for lawns and beds at a depth of 15–30 cm.

LR-CA

An increase in SOM was significantly associated with increased development age at a depth of 0–6 cm (F (1, 27) = 75.33, P = 2.5*10^–11, R²_adj = 0.61), but not at a depth of 6–30 cm (P = 0.86; Fig. 5b). Mean SOM increased by 3.4% (334% increase) between 0- and 20-year-old common areas.

Soil Carbon Accumulation Rates

For OTOW the estimated carbon accumulation at a depth of 0–15 cm was 0.156 and 0.064 kg C m-2 year-1 for lawns and landscape beds, respectively (Table 3). Changes in SOM were not significantly associated with increased lot age at a depth of 15–30 cm for OTOW, and thus a carbon accumulation rate was not estimated. For LR-R, an estimation of carbon sequestration at a depth of 0–15 cm yielded 0.084 and 0.100 kg C m-2 year-1, respectively, for lawns and landscape beds. Compared to surficial soil, carbon sequestration rates at a depth of 15–30 cm were four to five times less. For LR-CA, the change in mean SOM between 0- and 20-year-old common areas suggested a carbon accumulation rate of 0.060 kg C m-2 year-1, lower than other surface soils (OTOW & LR-R, 0–15 cm), but greater than LR-R rates at 15–30 cm.

Table 3

Soil carbon accumulation rates in lawns and landscape beds at depths of 0–15 and 15–30 cm. Changes in SOM were not significantly associated with lot age at a depth of 15–30 cm for OTOW, so a carbon accumulation rate was not estimated.
	Carbon accumulation rate (kg C m^-2 yr^-1)
Sample location & depth	OTOW	LR
Lawns 0–15 cm	0.156	0.084
Lawns 15–30 cm	NA	0.021
Beds 0–15 cm	0.064	0.100
Beds 15–30 cm	NA	0.020
Common Areas 0–6 cm		0.060
Common Areas 6–30 cm		NA

NA Not applicable.

Bulk density

Bulk densities over 1.80 g cm^-3 for sandy soils restrict root growth (USDA-NRCS, 2008). While bulk densities from deeper soils significantly decreased over time, they generally remained above or near 1.70 g cm^-3. We found a significant negative association between bulk density and lot age across all sampling depths, sampling sites, and sampling locations. In LR-R, bulk density decreased most rapidly with increasing lot age (0.024 g cm^-3 yr^-1) in lawns at a depth of 0–15 cm, and the least change in bulk density was observed in landscape beds (0.008 g cm^-3 yr^-1) a depth of 15–30 cm. In contrast, the relationship between bulk density and lot age was more consistent across sampling sites and depths for OTOW (0.013–0.014 g cm^-3 yr^-1). In LR-CA, changes in bulk density were similar between sampling depths, with mean decreases of 0.017 and 0.021 g cm^-3 yr^-1 at depths of 0–6 and 6–30 cm, respectively.

Past chronosequence studies have shown mixed results when analyzing the relationship between lot age and soil bulk density. Scharenbroch et al., (2005) found that in the Northwestern US, a sample of 48 urban and residential sites with a silt loam soil texture showed significantly reduced bulk density (0.006 g cm^-3 yr^-1, 18% difference) between 9- and 64-year-old sites. However, a 55-year chronosequence study by Raciti et al. (2011) of 32 residential sites in the Baltimore, Maryland, area found that lot age was not a significant predictor of bulk density, though residential bulk densities were not significantly different from forested soils.

The range of bulk densities observed in this study suggests a linear relationship to time. However, linear extrapolation is unsustainable; it is more likely that soils would reach an equilibrium density, a steady structural state. As vegetation is the main driver, though not the only contributor, one equilibrium point would be when porosity cannot be increased further without decreasing water or nutrient availability.

It is important to account for the initial condition or compaction level, in these types of studies when substituting space for time. By including landscapes from residential developments within the same communities, construction and landscape management methods are as uniform as possible over the lifetimes of these landscapes, therefore supporting the assumption that soil development processes are similar from lot to lot.

Soil Organic Matter

Amounts of SOM were similar for landscape beds and turfgrass for LR-R and OTOW, with higher accumulation at 0–15 cm than 15–30 cm. Significant increases in SOM were most prominent in surficial soils within depths of 0–15 and 0–6 cm across sampling sites. At a depth of 15–30 cm, a significant positive relationship between SOM and lot age was only observed in LR-R, and between 0- and 22-year-old lots, deeper SOM increased by a mean of 44%, compared to 160% increase for surficial soil depths of 0–15 cm. In LR-CA between 0- and 20-year-old, SOM increased by 336% within the surficial soil depth of 0–6 cm, which was the highest increase in SOM among study sites. There was very limited accumulation of SOM within LR-CA soils from the 15–30 cm depths. Managed urban turfgrass can accumulate higher levels of carbon compared to native soils (Pouyat et al., 2009).

Although SOM increased significantly (p < 0.05) over time within 15–30 cm depths under lawns and beds of LR-R, accumulation rates were much slower (0.02 kg C m^-2 yr^-1) than shallower soils. The accumulation rates within 15–30 cm depths were not significant for OTOW, nor for the 6–30 cm depth for LR-CA. Even after 20–30 years, SOM values are similar to that of new fill material. However, within the top 15 cm, SOM increased significantly (p < 0.001) under each vegetation coverage for both developments, with accumulation rates being greater in beds than lawns for LR-R and in lawns than beds for OTOW. Pooled mean for lawns and beds in LR-R was 0.092 kg C m^-2 yr^-1 and 0.110 kg C m^-2 yr^-1 for OTOW. This is likely due to soil being more compact at depth (15–30 cm) than near the surface (0–15 cm).

Soil Carbon Accumulation

Past chronosequence studies of carbon accumulation in urban soils of residential and recreational sites (i.e., golf courses) have found C accumulation rates between 0.026 and 0.355 kg C m^-2 yr^-1 at depths of up to 100 cm (Qian and Follett 2002; Raciti et al. 2011; Selhorst and Lal 2011; Huyler et al. 2014). Huyler et al. (2014) reported a relatively low carbon sequestration rate of 0.026 kg C m^-2 yr^-1 at a depth of 0–15 cm for pure lawns in Auburn, Alabama. The rates of carbon accumulation at 0–15 cm depth found in this study (0.064 to 0.156 kg C m^-2 yr^-1) are more akin to results of a study by Qian and Follett (2002), which found C accumulation rates of between 0.082 and 0.091 kg C m^-2 yr^-1 for soil samples collected at a depth of 0-11.4 cm in Midwestern US golf courses. Carbon accumulation in LR-CA turfgrass areas was less than that of LR-R and OTOW turfgrass sites, perhaps because of the differing turfgrass management strategies. It was surprising to find a relatively high C accumulation rate in our study given the preconception that coarse soil textures and warmer climates tend to relatively lessen C accumulation (Weil and Brady 2017). Managed turfgrass can accumulate higher level of carbon compared to native soils (Pouyat et al. 2009). For managed landscapes, this is likely more dependent on nitrogen inputs to landscapes from fertilization than from irrigation. Zirkle et al. (2022) estimated that humification of plant biomass accounted for 0.046–0.127 kg C m^-2 yr^-1, fertilization 0.078–0.098 kg C m^-2 yr^-1, and irrigation 0.005–0.010 kg C m^-2 yr^-1, for a total of 0.129–0.235 kg C m^-2 yr^-1 soil organic C accumulation. Thus, it is likely that fertilization and humification are the drivers of C accumulation in urban lawns, with mulch being the main source in landscape beds and mulched grass clippings in turfgrass areas.

The lack of SOC accumulation at depth (15–30 cm) seems counter to bulk densities decreasing over time and it does not reflect the increasing trend observed in 0–15 cm soil layers. The root zone for turfgrasses commonly grown in Florida is about 20 cm, with the majority of root mass within the top 15 cm. Horizontal distribution of root mass under lawns is more homogenous than within ornamental plant beds. As bulk densities increase with depth, it becomes more difficult for roots to penetrate deeper and deeper into the soil profile. Therefore, the tradeoff is between greater access to water and possibly nutrients vs. overcoming the soil structure.

Standardization of home building and land development practices suggests that these characteristics are representative of broader patterns in residential land use. While this study was conducted in Florida, urban landscapes are converging (Groffman et al. 2014). As land development in the United States moves towards planned communities, the approach is similar: to clear and grade the development site. The inferior quality of fill material and compaction creates a less than optimal growing condition for landscape vegetation. The underlying soil condition at construction sets the trajectory of the underlying soils for the future. However, turfgrass is the most common vegetation in many urban settings because it is the lowest cost option per unit area to purchase and install.

In addition, the annual cycle of root penetration and subsequent dieback is an important contributor to SOC accumulation. However, while bulk densities at 15–30 cm decreased over more than 20 years, the soil at this depth is still restrictive to turfgrass rooting. Further, frequent and shallow irrigation would promote shallower rooting depth, which may have also contributed to the lack of deeper SOC accumulation.

This is the first study to compare carbon accumulation rates under lawns to those under landscape beds. No distinct differences between lawns and landscape beds were found for bulk densities and soil organic matter. While soils underneath turfgrass accumulated soil carbon, these rates were similar to those within landscape beds. This suggests that equivalent or similar processes are occurring at similar rates within both parts of the landscape. From a management perspective, current irrigation and fertilization recommendations are lower for landscape beds than lawns in Florida. In both LR-R and OTOW, supplemental irrigation and fertilization are standard practices. As such, it can be concluded that landscape beds were likely more efficient for increasing SOC due to reduced inputs. This could have implications for efforts focused on increasing carbon sequestration and crediting.

To better understand the evolution of urban soil characteristics, future research should investigate whether the same processes are occurring within both lawns and beds, or whether similar but different processes are responsible for similar changes.

This study provides evidence for a positive relationship between soil quality and lot age in Central (OTOW) and Southwest Florida (LR) residential lawns and landscape beds with sandy soil textures, providing a framework for characterizing the evolution of residential soil characteristics in Florida and similar climatic and soil settings. Increased soil organic matter (0.060–0.156 kg C m^-2 yr^-1) and decreased bulk density (22% − 34%) were associated with increased lot age at a depth of 0–15 cm, and 0–6 cm for common areas. No broad differences were observed between the lawns and beds. While bulk densities at 15–30 cm decreased similarly to shallower soils, SOM accumulation was only significant at the LR sites, although the magnitude was negligible by comparison. This decoupling between SOM accumulation and decreasing bulk densities at depth is likely due to remnant compaction limiting that facilitate SOM accumulation, such as root penetration and cycling, and invertebrate activity. Estimated carbon sequestration rates were representative of past chronosequence studies that have largely been conducted in cooler climates with fine soil texture. This suggests that sandy soil texture and warm climate in Florida do not govern carbon sequestration rates as anticipated, but those conditions are overcome by vegetation type and landscape management practices (e.g., fertilization, irrigation, mulching, and mowing).

This study in combination with previous studies provides evidence that these processes likely occur over a much broader scale than the localized survey areas in this and previous studies. Future research should seek to better understand the spatiotemporal evolution of urban soils, driving factors, and the ecohydrological impacts. The costs and benefits of interventions, such as amending soil with compost via tillage, should be quantified to better inform policies pertaining to urban landscapes.

This is also the first known study to examine soil development characteristics under lawns and landscape beds. While turfgrass is recognized to carbon accumulation, this study provides evidence that landscape beds provide similar carbon sequestration, while typically requiring fewer inputs and maintenance. Better data and understanding of urban carbon cycles and evolution are needed to further understand the ecological states and trajectories of urban landscapes.

Acknowledgements

Financial support was provided by Southwest Florida Water Management District (SWFWMD) and the Florida Nursery, Growers, and Landscapers Association (FNGLA). Thank you to On Top of the World Communities and Lakewood Ranch communities and residents for allowing this research to take place. The staff from the Departments of Agricultural and Biological Engineering and the Soil, Water and Ecosystem Sciences at the University of Florida were instrumental in assisting with sample collection and analyses. We especially thank Manatee County UF/IFAS Extension and Michelle Atkinson for assisting with identifying residences in Lakewood Ranch.

Statements & Declarations

Funding

Southwest Florida Water Management District
Florida Nursery Growers & Landscapers Association

Competing Interests

No competing interests to declare.

Author Contributions

EB, AB, and HE conceived of the original project. AB and HE developed the research plan with support from EB. RF led data compilation and analyses and EB supervised the work, RF led development of the manuscript with support from EB. AB and HE contributed to the final manuscript. All authors provided critical feedback and helped shape the research, analysis, and manuscript.

Al-Shammary AAG, Kouzani AZ, Kaynak A, Khoo SY, Norton M, Gates W (2018) Soil Bulk Density Estimation Methods: A Review. Pedosphere 28:581–596. https://doi.org/10.1016/S1002-0160(18)60034-7
Box GEP, Cox DR (1964) An Analysis of Transformations. J R Stat Soc Ser B 26:211–252
Carr MH, Zwick PD (2016) TECHNICAL REPORT: Florida 2070: Mapping Florida’s Future – Alternative Patterns of Development in 2070. https://1000friendsofflorida.org/florida2070/wp-content/uploads/2016/09/florida2070technicalreportfinal.pdf
Chen Y, Day SD, Wick AF, McGuire KJ (2014) Influence of urban land development and subsequent soil rehabilitation on soil aggregates, carbon, and hydraulic conductivity. Sci Total Environ 494–495:329–336. https://doi.org/10.1016/J.SCITOTENV.2014.06.099
Cogger CG (2005) Potential compost benefits for restoration of soils disturbed by urban development. Compost Sci Util 13. https://doi.org/10.1080/1065657X.2005.10702248
Dieter CA, Maupin MA, Caldwell RR, Harris MA, Ivahnenko TI, Lovelace JK, Barber NL, Linsey KS (2018) Estimated use of water in the United States in 2015. U S Geological Survey Circular 1441. https://doi.org/10.3133/cir1441. 65 p
Gee GW, Bauder JW (1986) Particle size analysis. In: Klute A (ed) Methods of soil analysis, part 1. Physical and mineralogical methods, 2nd edn. American Society of Agronomy, Madison (WI), pp 383–411
Groffman PM, Cavender-Bares J, Bettez ND, Grove JM, Hall SJ, Heffernan JB, Hobbie SE, Larson KL, Morse JL, Neill C, Nelson K, O'Neil-Dunne J, Ogden L, Pataki DE, Polsky C, Chowdhury RR, Steele M (2014) Ecological homogenization of urban USA. Front Ecol Environ 12:74–81. https://doi.org/10.1890/120374
Gregory JH, Dukes MD, Jones PH, Miller GL (2006) Effect of urban soil compaction on infiltration rate. J. Soil Water Conserv. May 2006, 61 (3) 117–124
Florida State University (2023) Florida statewide averaged precipitation data (1895 to 2021): Tallahassee, Fla., Florida Climate Center, Office of the State Climatologist, accessed February 25, 2023, at https://climatecenter.fsu.edu/products-services/data/statewide-averages/precipitation
Hudson BD (1994) Soil organic matter and available water capacity. J Soil Water Conserv 49:189–194. https://doi.org/10.1201/9780429445552-36
Huyler A, Chappelka AH, Prior SA, Somers GL (2014) Drivers of soil carbon in residential “pure lawns. in Auburn Alabama Urban Ecosyst 17:205–219. https://doi.org/10.1007/s11252-013-0294-3
Jim CY (1998) Urban soil characteristics and limitations for landscape planting in Hong Kong. Landsc Urban Plan 40:235–249. https://doi.org/10.1016/S0169-2046(97)00117-5
Karlen DL, Mausbach MJ, Doran JW, Cline RG, Harris RF, Schuman GE (1997) Soil quality: a concept, definition, and framework for evaluation (a guest editorial). Soil Sci Soc Am J 61:4–10. https://doi.org/10.2136/SSSAJ1997.03615995006100010001X
Kozlowski TT (1999) Soil compaction and growth of woody plants. Scand J For Res 14:596–619. https://doi.org/https://doi.org/10.1080/02827589908540825
Kumar Sen P (1968) Estimates of the regression coefficient based on kendall’s tau. J Am Stat Assoc 63:1379–1389. https://doi.org/https://doi.org/10.2307/2285891
Manatee Country Property Appraiser (2021) Statistics and Owner Information – Manatee County Property Appraiser. https://www.manateepao.com/statistics/. Accessed 2 Jul 2021
Marella RL (2020) Water withdrawals, uses, and trends in Florida, 2015: U.S. Geological Survey Scientific Investigations Report 2019–5147, 52 p., https://doi.org/10.3133/sir20195147
Marion County Property Appraiser (2022) MCPA Property Search. https://www.pa.marion.fl.us/PropertySearch.aspx. Accessed 18 May 2022
Milesi C, Running SW, Elvidge CD, Deitz JB, Benjamin TT, Nemani RR (2005) Mapping and modeling the biogeochemical cycling of turf grasses in the United States. Environ Manage 36:426–438. https://doi.org/10.1007/s00267-004-0316-2
NOAA (2021) NOAA Global Hourly - Integrated Surface Database (ISD). In: Natl. Centers Environ. Inf. https://www.ncei.noaa.gov/products/land-based-station/integrated-surface-database. Accessed 17 Oct 2021
NRCS (2007) Florida Soil Characterization Data Retrieval System. United States Department of Agriculture-Natural Resources Conservation Service & Florida Department of Transportation. Retrieved from http://soils.ifas.ufl.edu/flsoils/
Pachon JC, Kowalski KR, Butterick JK, Bacon AR (2019) Quantified effects of particle refractive index assumptions on laser diffraction analyses of selected soils. Soil Sci Soc Am J 83(3):518–530
Pitt R, Lantrip J, Harrison R, Henry CL, Xue D (1999) Infiltration Through Disturbed Urban Soils and Compost- Amended Soil Effects on Runoff Quality and Quantity. United State Environmentatl Protection Agency National Risk Management Research Laboratory. EPA/600/R-00/016. https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NRMRL&dirEntryId=64168
Pouyat RV, Yesilonis ID, Golubiewski NE (2009) A comparison of soil organic carbon stocks between residential turf grass and native soil. Urban Ecosyst 12:45–62. https://doi.org/10.1007/s11252-008-0059-6
Pribyl DW (2010) A critical review of the conventional SOC to SOM conversion factor. Geoderma 156:75–83. https://doi.org/10.1016/J.GEODERMA.2010.02.003
Qian Y, Follett RF (2002) Assessing Soil Carbon Sequestration in Turfgrass Systems Using Long-Term Soil Testing Data. Agron J 94:930–935. https://doi.org/10.2134/AGRONJ2002.9300
Raciti SM, Groffman PM, Jenkins JC, Pouyat RV, Fahey TJ, Picket STA, Cadenasso ML (2011) Accumulation of carbon and nitrogen in residential soils with different land-use histories. Ecosyst 2011 142 14:287–297. https://doi.org/10.1007/S10021-010-9409-3
Ryan CD, Groffman PM, Grove JM, Hall SJ, Heffernan JB, Hobbie SE, Locke DH, Morse JL, Neill C, Nelson KC, O'Neil-Dunne J, Chowhury RR, Steele M, Trammell TLE (2022) Ecological homogenization of soil properties in the American residential macrosystem. Ecosphere 13(9):e4208. https://doi.org/10.1002/ecs2.4208
Scharenbroch BC, Lloyd JE, Johnson-Maynard JL (2005) Distinguishing urban soils with physical, chemical, and biological properties. Pedobiologia 49(4):283–296. https://doi.org/10.1016/j.pedobi.2004.12.002
Selhorst AL, Lal R (2011) Carbon budgeting in golf course soils of Central Ohio. Urban Ecosyst 14:771–781. https://doi.org/10.1007/S11252-011-0168-5/TABLES/4
Web Soil Survey. In: Nat. Resour. Conserv. Serv. - United States Dep. Soil Survey Staff, Agric (2021) http://websoilsurvey.sc.egov.usda.gov/
U.S. Census Bureau (2023) QuickFacts Florida, Population, Census, April 1, 2020: accessed February 25, 2023, at https://www.census.gov/quickfacts/FL
USDA-NRCS (2022) Myakka -- Florida State Soil. https://www.nrcs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb1236953.pdf. Accessed 8 Jun 2022
USDA-NRCS (2008) "Soil Quality: Indicators: Bulk Density" Soil Quality for Environmental Health. http://www.soilquality.org/indicators/bulk_density.html. Accessed: 11 Jan 2023
USDA (1987) Soil Mechanics Level I, Module 3 - USDA Textural Soil Classification Study Guide. https://www.wcc.nrcs.usda.gov/ftpref/wntsc/H&H/training/soilsOther/soil-USDA-textural-class.pdf. Accessed 2 Jul 2021
Visit Florida (2023) Florida-fact, 2019: Accessed February 25, 2023, at https://www.visitflorida.org/resources/research/
Weil RR, Brady NC (2017) The Nature and Properties of Soils, 15th edn. Pearson Education Limited

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Development of Sandy Soil Properties Over Time within Subtropical Residential Landscapes

Status:

Version 1

Abstract

Figures

Introduction

Materials and Methods

Site Descriptions

Soil Sampling

Soil Sample Analyses

Bulk Density

Soil Texture

Soil organic matter

Data Analysis

Results

Soil Particle Size

Bulk Density

OTOW

LR-R

LR-CA

Soil Organic Matter

OTOW

LR-R

LR-CA

Soil Carbon Accumulation Rates

Discussion

Bulk density

Soil Organic Matter

Soil Carbon Accumulation

Conclusions

Declarations

References

Additional Declarations

Status:

Version 1