Determining Variability in Characteristics of Residential Landscape Soils that Influences Infiltration Rates

Soil Textural Classification

The proportion of sand within lots ranged from 81.9% to 97.8% (data not shown), while the mean proportion of sand within communities ranged from 91.0% to 96.4 %. Sand particle size fractions dominated soils (Table 1). Textural analysis determined 90% of soils (36 lots) were classified as sand, with only three lots containing loamy sand and one lot containing a sandy loam. Although the proportion of sand varied between samples, no difference in the total percent of sand could be found. Similar percentages of soil sand (89.3% to 96.2%) were observed in a related study in north-central Florida (Gregory et al. 2006) and are typical within central Florida (Brown et al. 1990). Medium (0.25 to 0.5 mm) and fine (0.1 to 0.25 mm) sand particle sizes dominated the sand fraction, with means of 386.2 and 455.1 g kg⁻¹ soil, respectively (Table 1) (USDA 1995). Coarse sand particle (0.5 to 1.0 mm) size fractions were minimal and ranged from 3.2 to 64.7 g kg⁻¹ soil.

NRCS soil maps indicated that the sampled areas were mapped as Spodosols, Entisols, and Ultisols prior to urban residential development (NRCS 2011). These soil orders are common throughout Florida and the southeastern U.S. region (Brady and Weil 2002). Results of textural classification analysis were similar to soil descriptions listed in NRCS soil maps for most (90%) sampled locations. In lots where soil textural classification differed from the NRCS soil map description, clay content of soil was consistently higher. This most likely was due to the use of non-native soil in site leveling and grading activities (C.R. Glagola, pers. comm.).

Pore Size Distribution

Total soil porosity ranged from 38.8 to 46.8 cm³ 100 cm⁻³ soil (Table 2). Mean macropore and micropore soil volume was 9.6 and 24.7 cm³ 100 cm⁻³ soil, respectively. Micropores (diameter 0.05–0.0002 mm) were the dominant soil pore size among sampled lots, representing 58% of total mean soil pore space. Macropores (diameter > 0.05 mm) and cryptopores (diameter < 0.0002 mm) represented 22% and 20% of total mean soil pore space, respectively. The coefficient of variation for mean macroporosity within each lot ranged from 1.2% to 123.7%, whereas variation within communities ranged from 35.7% to 85.1%. High variability in macroporosity between replicate samples was likely a result of lithologic discontinuities resulting from construction activities (Craul 1985). Dominance of soil micropores restrict air and water movement (Tuli and Hopmans 2004; Tuli et al. 2005). This condition reduces infiltration rates and promotes urban stormwater “fiashiness” during intense precipitation events (Ehrenfeld et al. 2003; Baker et al. 2004). These conditions may be of concern in residential landscape soils where runoff conditions may contribute to non-point source pollution (Brett et al. 2005; Atasoy et al. 2006).

Soil Moisture Retention Characteristics

A soil moisture characteristic curve representing the mean moisture retention volumes of all collected samples is shown in Figure 1. Approximately 42% of the total mean volumetric water content was held as gravitational water (0 to 10 kPa), while approximately 66% of the total mean volumetric water content was held between 0 and 100 kPa. The large percentage of water held at low pressure potentials was expected, given soils were dominated by sand particle size fractions. Mean soil moisture retention volumes of communities ranged from 16.5 to 28.5 mL water 100 cm⁻³ soil between moisture potentials 0 and 100 kPa (Table 3). Significant differences in soil moisture content were observed among communities. The coefficient of variation for mean soil moisture content within communities ranged from 16.7% to 50.8% (Table 3). High variability between replicate samples was likely a result of heterogeneity in soil pore size and its correlated influence on soil moisture retention (Craul 1985; Jeetendra and Singh 2008). Low soil moisture retention values were a result of the high percentage of coarse, sandy textured soils (Goncalves et al. 2010). Soil moisture retention characteristics of urban residential soils have not previously been reported.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Mean soil water characteristic curve for all samples collected within nine newly established urban residential communities within central Florida. Error bars represent the standard error of the mean.

Soil Infiltration

Mean soil infiltration rates in communities ranged from 11.2 to 63.6 cm hr⁻¹. Significant differences were observed among communities (Table 3). High variability in soil infiltration rate was also observed within communities. The maximum recorded infiltration rate was 111.1 cm hr⁻¹, while the minimum rate was 2.0 cm hr⁻¹. Similar infiltration rates and variability has been observed in related studies (0.3 to 220.0 cm hr⁻¹, Pitt et al. 1999; 0.8 to 17.5 cm hr⁻¹, Gregory et al. 2006).

Florida stormwater modeling and planning regulations require that management plans account for potential stormwater flow. Infiltration rates in 45% of sampled lots were below the 100-year, 24-hour design storm intensity of 26.7 cm hr⁻¹ for Orlando, Florida (FDEP 2006). Further, 30% of sampled lots had infiltration rates below the 5-year, 24-hour design storm intensity of 16.5 cm hr⁻¹. Thus, a large percentage of sampled lots would likely produce stormwater runoff under intense precipitation events. Penetration of landscape plant material roots into compacted urban soil may increase infiltration. Bartens et al. (2008) observed a 63% increase in mean infiltration rate as a result of black oak (Quercus velutina Lam.) and red maple (Acer rubrum L.) root penetration into compacted container soil. Authors attributed this observation to the formation of water channels along live root channels. Additional research examining the effect of plant material root growth on soil compaction and infiltration rate is needed.

Soil Bulk Density

Mean soil D_b values ranged from 1.65 to 1.78 g cm⁻³ (Table 3). These values were similar to those observed by Gregory et al. (2006), where relationships between residential construction equipment use and soil compaction in northern central Florida were examined. Observed D_b values were also similar to those found in newly constructed urban residential communities within Idaho and Washington states (Scharenbroch et al. 2005). Over 50% of sampled communities had mean D_b values of approximately 1.75 g cm⁻³. When compaction density exceeds this value in predominately sandy soils, plant root soil penetration can be negatively impacted (Daddow and Warrington 1983; Voorhees 1992). Ensuring soil compaction remains below this threshold may be necessary when attempting to successfully transplant and establish plant materials in similar, newly constructed residential landscapes. Delineation of construction traffic zones are recommended to minimize unintended soil compaction (Randrup and Dralle 1997). Post-construction mechanical loosening of compacted soil and addition of organic amendments are recommended to decrease D_b and increase soil macroporosity (Rivenshield and Bassuk 2007).

Soil compaction in newly constructed urban residential communities within central Florida had low variability at both small (lot) and large (community) spatial scales and was likely the result of similar, unintentional soil compaction that resulted from site leveling and grading practices. Variability of D_b within each sampled lot was low. Coefficient of variation of D_b within communities ranged from 3.3% to 7.6% (Table 3). Several individual lots were observed to have significantly greater D_b values than other lots within the same community (data not shown). This trend was random and likely a result of additional site traffic or equipment storage on these lots. Visual observation was made of construction equipment and vehicular storage on random lots at the time of sampling. Efforts to avoid these lots during sample collection were made; however, it was not possible to determine if sampled lots had additional traffic-induced compaction prior to sample collection.

Although no housing structures were erected at the time of sampling, significant increases in soil compaction resulting from additional construction activities are not likely given the high soil compaction values already imposed. Assuming no additional soil disturbances occur, temporal decreases in D_b are likely to occur as a result of improvements in physical, chemical, and biological soil characteristics (Scharenbroch et al. 2005).

Despite examining relationships between soil infiltration and soil moisture volume at various pressure deficits iteratively, no correlations were observed. Similarly, soil bulk density was not a significant predictor of infiltration rate here, although its relationship with infiltration is well established (Grabosky and Bassuk 1995; Jim 1998; Gregory et al. 2006; Pitt et al. 2008). Consistent construction practices likely caused uniformity among D_b measurements, even though the study encompassed sites managed by various contractors. Regression analysis determined percentage total sand was the only significantly correlated variable to infiltration rate; however, percentage silt and clay were added to the model given their relationship to soil texture. Development of the regression model allowed for calculation of predicted infiltration rate. Relationships between recorded and predicted infiltration rate were examined (Figure 2). One outlier was omitted given its predicted value was >2 standard deviations from the recorded value. The coefficient of determination for the regression model was low (r² = 0.43) and likely a result of high variability among recorded infiltration values.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Regression of recorded infiltration rate (dots) and predicted infiltration rate of soil within nine newly established urban residential communities in central Florida. Predicted infiltration rate was calculated through the regression model: Infiltration predicted = −1497.09 + 16.07 (% sand content) + 6.58 (% silt content) + 13.49 (% clay content).