Fill Soil Effects on Soil Aeration and Tree Growth

METHODS

In 1996, 3 years prior to the installation of experimental treatments, a field plot [13 × 26 m (43 × 87 ft)] was excavated to a depth of 30 cm (12 in.) at the Armstrong Field Research Area on the campus of the University of California, Davis, U.S. (Figure 1). This “sunken plot” design was used to create a condition where 30 cm (12 in.) of fill soil could be added back around established trees so that the top of the fill would be even with original grade. This design was chosen so that there would be no nearby “edge” to the fill facilitating horizontal gas diffusion. The large area planned for fill, combined with the surrounding soil mass at the same elevation as the finished fill, was used to maximize any potential effects of the fill on trees.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Cherry trees (Prunus × yedoensis) were planted in a field plot that was excavated 30 cm (12 in.) deep prior to planting. After a 3-year establishment period, fill soil (30 cm) was added to two subplots (fill-only and fill plus aeration treatments), while one subplot remained without fill (control).

Laboratory analysis of plot soil by the ANR Analytical Laboratory (University of California, Davis) determined that pH = 7.1 and electrical conductivity (ECe) = 0.43 dS/m. Soil texture was classified as loam (44% sand, 38% silt, and 18% clay) at 15 cm (6 in.) below the excavated grade, and sandy loam (56% sand, 31% silt, and 13% clay) at 45 cm (18 in.) below the excavated grade. Bulk density in the base soil was determined from three samples taken at each of two depths in each of three subplots. Values across subplots ranged from 1.62 to 1.74 g/cm³ at 15 cm (6 in.), and 1.31 to 1.42 g/cm³ at 45 cm (18 in.) (Figure 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Soil bulk density (g/cm³) at two depths [15 and 45 cm (6 and 18 in.)] in each of the treatment subplots.

Following excavation, 45 cherry trees [Prunus × yedoensis ‘Afterglow’, 19 L (5 gal) container stock], were planted in the excavated plot. Cherry was selected because of its reported susceptibility to flooding (oxygen stress) injury (Rowe and Beardsell 1973). Five trees were spaced 2.3 m (7.7 ft) apart in each of nine rows that were spaced 2.3 m between center lines. All trees were irrigated after planting and at regular intervals thereafter to establish a network of roots throughout the plot. The plot surface was sloped (approximately 2%) to one end where surface water (from rainfall and irrigation) drained into a collection pit and could be pumped out to avoid standing water. Trees were maintained for three growing seasons before fill was installed.

In 1999, after a 3-year establishment period, the main plot was divided into three subplots consisting of three adjoining rows of trees with five trees in a row. One of the following treatments was assigned to each subplot: (1) fill soil only, (2) fill soil plus an aeration system, and (3) no fill soil (control). Header boards [30 cm (12 in.) high] spanned the plot, separating the unfilled control from the adjoining fill treatment. Soil removed during pit excavation was reused as the fill soil.

In both fill soil treatments, soil was added in two lifts of 15 cm (6 in.) each. A gasoline engine—powered percussion rammer and a hand-tamping tool were used to compact each lift to a bulk density of 1.6 g/cm³. In the fill plus aeration treatment, a continuous line of flexible, perforated drain pipe [10 cm (4 in.) diameter] was placed on the soil surface before any fill was added. The piping formed eight rows [11 m (36 ft) lengths] and was ventilated to the aboveground atmosphere using vertical vents positioned at the midpoint and the ends of the piping system.

Prior to fill installation, a 45 cm (18 in.) length of flexible plastic pipe [10 cm (4 in.) diameter] was placed around the lower trunk of all trees to avoid direct contact between fill soil and trunk tissues.All plots were irrigated immediately after fill soil installation to bring soil moisture content to field capacity. A microsprinkler irrigation system was installed on the surface to supply water to the plot for the duration of the study.

Measurements

One month prior to fill installation, oxygen diffusion rates and matric tensions of the soil were measured for 11 days within each subplot to establish a pre-treatment baseline. Oxygen diffusion rate was measured using platinum-tipped microelectrodes connected to an ODR ratemeter (Jensen Instruments, Tacoma, WA) as described by Stolzy and Letey (1964). Clusters of five electrodes were used at each of three locations in each subplot. Electrodes were inserted to depths of 15 and 45 cm (6 and 18 in.) in the base soil (Figure 3). The total array of 15 electrodes at each depth was used to calculate the mean ODR reading at that depth for each measurement interval in each subplot. This procedure allowed a direct comparison of root zone ODR before and after fill application. Electrodes were removed prior to fill installation and then reinstalled following installation. After fill installation, ODR measurements were repeated two times, with each measurement period spanning 14 days. For a 10-day period between the two measurement periods, all electrodes were removed, cleaned, and then reinstalled.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

ODR measurements were taken at 15 and 45 cm (6 and 18 in.) depths in the base soil of each subplot.

Soil moisture was measured using tensiometers (Soil Moisture Equipment Corp., Santa Barbara, CA) placed adjacent to clusters of ODR electrodes in each subplot at 15 and 45 cm (6 and 18 in.) depths in the base soil. Previous studies found a negative relationship between ODR and soil moisture: As soil moisture increased, ODR declined (MacDonald et al. 1993; Costello et al. 1994).

To assess the impact of fill treatments on tree water status, midday stem water potential was measured before and after irrigation in summer 2000, approximately 1 year after fill addition. Four leaves from each of three trees in each subplot were removed, and water potential was measured using a pressure bomb. Following methods described by Shackel et al. (1998), leaves were enclosed in plastic pouches for 3 hours prior to removal and measurement.

Following stem water potential measurements, the root zone around trees was exposed using a pneumatic excavation tool (AirSpade®). In fill subplots, fill soil was initially excavated to the original grade (i.e., to the surface of the base soil). From this entry level, further excavation in all subplots exposed roots to a depth of 45 cm (18 in.). Although root mass was not measured, observations were made of root health and distribution.

Trunk diameter, stem water potential, ODR, and soil moisture data were analyzed using Fisher’s protected LSD test on a one-way randomized ANOVA design.

RESULTS

Plant Response

All trees survived and grew normally over the year following fill installation. Trunk diameter growth was not negatively impacted by fill treatments. Growth in the fill plus aeration treatment was significantly greater than that of controls and the fill-only treatments, while trunk diameter growth in the fill-only treatment was equivalent to controls (Figure 4).

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Trunk diameter growth (1-year means) for three fill treatments. Means with no letter in common are significantly different using Fisher’s protected LSD test (P = 0.01).

Although shoot growth was measured, results are not reported because growth was found to be highly variable throughout the canopy due to differences in shoot position and exposure.

In terms of visual appearance, leaf color and canopy density were similar across all treatments (Figure 5). There was no leaf drop, chlorosis, or dieback noted in any trees. Similar results were obtained in previous experiments evaluating fill effects on long-established cherry (Prunus mahaleb) trees (Tusler et al. 1998).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

One year after fill treatments, all trees appeared equivalent in canopy density and leaf color. Trees in the foreground (in front of header board) are in the control subplot, while those in background are in the fill-only and fill plus aeration subplots. No signs of leaf chlorosis, leaf drop, or branch dieback were evident on any of the trees.

Before irrigation, when soil moisture tensions were 40 to 55 centibars (cb), midday stem water potential was highest (most negative) in fill-only trees (−1.91 MPa) and lower in fill plus aeration (−1.60 MPa) and control (−1.64 MPa) trees (Figure 6). After irrigation, stem water potential declined in all treatments but remained highest in fill-only trees (−1.31 MPa) and similar in fill plus aeration (−1.17 MPa) and control (−1.12 MPa) trees. Measurements prior to irrigation suggest a moderate level of water stress (−1.5 to −2.0 MPa), while those after irrigation suggest a mild level of water stress (−1.0 to −1.5 MPa) (Shackel et al. 1998).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Midday stem water potential before and after irrigation for all treatments. Measurements were taken 1 year after fill treatments were installed. Means with no letter in common are significantly different using Fisher’s protected LSD test (P = 0.01).

Upon excavation, roots were found both in the original-grade and fill soils. In the original-grade soil, shallow roots developed laterally above a subsurface compact layer, while other roots developed vertically into the deeper sandy loam layer. Roots appeared healthy and uninjured by the addition of fill. Notably, a concentration of roots at the interface between the fill and original-grade soils was not found. We did, however, observe an abundance of roots in the fill soil. Most had a vertical orientation, growing toward the soil surface, where they branched extensively. In the fill plus aeration treatment, roots did not develop preferentially around aeration pipes. Instead, they grew upward past the pipes toward the soil surface and then branched near the soil surface (Figure 7).

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Roots in the fill plus aeration subplot grew from the base soil into the fill, past the aeration pipe, and branched near the fill surface. Roots were painted white for contrast.

During excavations, guards placed around tree trunks were removed and trunk tissues inspected. Bark was dry and appeared healthy: It was not discolored, water-soaked, or cracked.

DISCUSSION

While many studies have attempted to characterize the impact of fill soil on root zone aeration and tree performance, the data have been difficult to interpret because there have been confounding factors. For example, in landscape fills, there is a considerable amount of variability between surface area and side area: Shallow fills have little side area, while deep fills can have substantial amounts. Relative to surface area, this difference in depth was considered to have an impact on air relations in the fill. In addition, fill may be bordered by hardscape (such as a retaining wall) or feathered out to match original grade. Both cases increase the potential for gas intrusion from the side. A similar assessment regarding side effects was made by Smith et al. (1995). Our subgrade plot design was used to minimize side effects and limit associated variability.

In our experiments, fill did not have a negative impact on tree growth or health. Growth in fill subplots was equivalent to or greater than that for trees without fill (controls). This finding is consistent with a previous study evaluating fill effects on 6-year-old cherry (Prunus mahaleb) trees (Tusler et al. 1998). In addition, Day et al. (2001) reported similar results applying 20 cm (8 in.) of sandy loam fill (compacted and noncompacted) over the root zone of 22-year-old white oak (Quercus alba) and 13-year-old sweetgum (Liquidambar styraciflua). After 3 years, no consistent effect of fill on growth (trunk diameter or height), chlorophyll fluorescence, or soil respiration was found in either species.

It is not clear why trunk diameter growth was significantly greater in the fill plus aeration subplot than in fill-only or the control treatments. This result probably is not related to soil aeration differences, however, because ODR levels in both fill treatments were not significantly different. If the aeration system had enhanced aeration levels, then a difference in ODR levels between the two treatments should have been found. Additionally, roots in the fill plus aeration subplots were observed to have grown past aeration pipes toward the surface of the fill. Root branching and proliferation occurred principally near the fill surface, not next to aeration pipes. If the aeration piping had actually enhanced oxygen levels in the fill soil, roots should have been more abundant next to the pipes. Collectively, these findings suggest that aeration enhancement was not a factor contributing to increased trunk diameter in the fill plus aeration subplot. They also indicate that the aeration piping had no detectable beneficial effect on soil ODR or root response. Similar results were reported by Tusler et al. (1998), in which a subsurface aeration system did not increase ODR level in underlying field soil or in fill. In addition, previous laboratory and greenhouse studies indicated that increases in ODR levels were not found in the soil adjacent to piping systems or core vents (MacDonald et al. 1994; Tusler 1999).

Differences in trunk growth were more likely related to differences in water relations. In the fill-only subplot, copious amounts of water were needed to sufficiently wet the underlying field soil. Water was applied for many hours to generate soil moisture tension readings equivalent to those found in the fill plus aeration and the control subplots. As a result of this differential wetting pattern, trees in the fill-only treatment may have experienced periods of water stress of uncertain duration. In fact, midday stem water potential measurements indicated that water stress levels were higher in trees in the fill-only treatment than in the fill plus aeration treatment or the control treatment (Figure 6). If fill-only trees did experience episodes of water stress, then growth would be expected to be less than that found in unstressed trees.

Similarly, Day et al. (2001) reported that fill disrupted normal soil moisture patterns in both white oak (Quercus alba) and sweetgum (Liquidambar styraciflua) plots. In both cases, soil underlying fill was drier than the fill layer. They noted that changes in water movement through the soil profile associated with fill may have a long-term effect on tree growth.

Although differences in water relations may explain the growth differential between fill-only and fill plus aeration treatments, water stress probably was not the cause of the trunk growth difference between fill plus aeration trees and controls: Water infiltration into the control subplot was not impeded by fill. It is possible that extensive root development into the fill soil (found in both fill treatments) caused a growth enhancement effect relative to the control treatment, however. Although not measured, the mass of roots found in fill soil suggests that fill treatments may have provided a greater volume of soil favorable for root development than that in the control subplot. With this greater soil volume, trees in fill subplots may have had a higher potential for growth. This growth potential differential may not have been expressed in the fill-only treatment, however, due to water stress episodes (as discussed above).

The lack of reduction in ODR resulting from fill may be due in part to initially low ODR levels at the 15 cm (6 in.) depth. For all measurement periods, ODR at 15 cm was lower than the level considered critical for root function (0.2 mg/cm²/min) (Stolzy and Letey 1964). Bulk density in this zone ranged from 1.62 to 1.74 g/cm³, which is greater than the bulk density range (1.45 to 1.60 g/cm³) considered to be growth-limiting for loam (Daddow and Warrington 1983). At this bulk density, low ODR values likely resulted from highly limited pore space for gas diffusion. If so, then the potential for fill to further depress ODR level would have been low. In the 45 cm (18 in.) zone, however, ODR levels were substantially higher than those at 15 cm (6 in.), which were likely related to a lower bulk density (1.31 to 1.42 g/cm³), well below the growth-limiting level for sandy loam (1.60 to 1.75 g/cm³). Nonetheless, even though a greater potential for a reduction in ODR existed, there was only one measurement period (out of four) in which ODR levels in fill treatments were less than controls. Ostensibly, soil physical conditions (bulk density and porosity) were sufficient to avoid a fill-induced depression of soil aeration.

Even though fill soil did not cause a consistent or substantial impact on soil aeration, as measured by ODR, or plant growth, this finding should not be interpreted to mean that fill soils cannot depress root zone ODR levels or cause plant injury. Rather, it is only under the established experimental conditions that aeration was not reduced to levels that resulted in plant injury. It is certainly possible that fill soil depth, extent, texture, density, or porosity simply did not reach a critical level for plant injury in these studies. Our efforts to compact the fill soil and to prevent horizontal air movement through exposed edges were designed to create a severe fill event, however.

Finally, tree guards used to protect trunk tissues from contact with fill soil may have played an important role in minimizing injury. Without some protection for trunk tissues, fill potentially could cause a negative impact on growth and health. Harris et al. (1999) emphasize the need for the protection of trunk tissues from contact with fill soils not only to minimize disease development but also to avoid the potential of structural instability caused by adventitious root development. However, Day et al. (2001) reported neither a beneficial nor detrimental effect of tree wells on the growth or health of white oak (Quercus alba) and sweetgum (Liquidambar styraciflua). This assessment was made 3 years after the application of fill, however, and a longer time period may needed for the expression of adverse health effects.

To minimize the potential for plant injury resulting from a fill event, all contributing factors should be considered. These include pre-fill factors such as soil compaction and mechanical injury to roots, and post-fill factors such as water stress, aeration deficit, and disease resulting from fill soil contact with trunk tissues. Day et al. (2001) emphasize that “other factors associated with raising the grade, such as soil trafficking and root severance, may be responsible for much of the tree decline attributed to fill.” Accordingly, the level of plant injury associated with fill will likely range from inconsequential to severe depending on pre-fill site conditions, fill installation practices, and post-fill management practices.