Above and Belowground Growth of Corymbia maculata in a Constructed Soil: The Effect of Profile Design and Organic Amendment

Experimental Profiles

Three types of organic matter, coir fiber, composted green waste and composted biosolids were evaluated. Each was used as a soil amendment at the rates of 5, 10, and 20% by volume. Coir fiber is a product of the husk of the coconut, and is also known as coir fiber dust. It is used as a peat replacement in nursery media (Offord et al. 1998) and has been used in constructed landscape soils. The composted green waste was commercially composted to meet the criteria of Australian Standard 4434-2003 Mulches, composts and soil conditioners (Standards Australia 2003b). The composted biosolids met the Class A requirements of the United States Environment Protection Authority (1993) biosolids standards, and the AS 4454-2003 requirements for pasteurization. Both the composted green waste and composted biosolids were passed through a 5 mm (0.2 in) screen before use.

The other factor was profile design where the three organic amendments were either restricted to the surface 150 mm (6 in) of the profile (layered profile) or mixed through the entire profile (uniform profile). In each design the organic amendment of the treated soil was at the three rates given above. The soil mixes were prepared in 50 L (13 gal) batches in a concrete mixer. Trace elements (0.8 kg m⁻³ Micromax, Scott’s Australia, Baulkham Hills, NSW) and ground dolomite (1 kg m⁻³) (0.5 and 0.6 lb ft⁻³ respectively) were mixed in at this stage, prior to profile construction.

Profile Construction

A 25 mm layer of 4–5 mm angular gravel was placed in the bottom of each container as a filter and drainage layer. The sand or sand/organic mixes were placed in the containers in 150 mm layers, and each layer was lightly compressed manually. The total depth of root zone was 600 mm (24 in), with a 25 mm deep irrigation volume left at the top of all planters to facilitate irrigation by micro-sprays. After filling, the planters were irrigated several times to fully consolidate the constructed profile prior to planting. The bulk density of the constructed profiles ranged from 1.9 Mg m⁻³ (3200 lb yd⁻³) for the unamended sand to 1.5 Mg m⁻³ (2530 lb yd⁻³) for sands amended with 20% organic matter.

Planting Stock, Planting and Maintenance

Five-month old C. maculata seedling liners were purchased in square containers [50 mm wide x 130 mm (2 in x 5 in) deep]. They were selected for the experiment based on their uniformity of root system development, size, and leaf morphology. Relatively small plants were used because of the restricted volume of the experimental system. A single nutrition treatment was surface applied to all constructed profiles after planting. The slow release fertilizer Once High Fine N (11:1:4) (Scott’s Australia) was used. This fertilizer was reapplied four months later at the same rate giving total N, P and K applications the equivalent of 1000, 90, and 340 kg ha⁻¹ (890, 80, and 300 lb ac⁻¹) respectively. These rates were based on growth responses observed in an initial experiment that evaluated nutrient responses in C. maculata growing in a sand profile with the surface 150 mm amended with coir fiber (Smith 2003). In greenhouse experiments to evaluate nutrient responses in forestry eucalypts, Olsen and Bell (1990) used an upper rate of 200 kg N kg ha⁻¹ (180 lb N ac⁻¹) in a clay loam soil, and Bennett et al. (1996) an upper rate of 400 kg N ha⁻¹ (360 lb N ac⁻¹) in a clay soil. The higher fertilizer used in this experiment was due to the low fertility of the sandy soil. Initial fertilizer recommendations for plantation eucalypts planted in sandy soil in western Australia are as high as 18 g N (0.63 oz N) per tree (Cromer 1996) compared to the 3.9 g N (0.14 g N) per tree applied in this experiment. The experiment was conducted over the spring and early summer with the trees planted in the containers on August 2, 2000, and harvested on January 2, 2001. During the experiment the trees were irrigated with a microirrigation system controlled with a timer, as required to maintain soil moisture. All blocks and containers received the same irrigation treatment. The trees were grown outdoors in a nursery in Richmond, VIC, Australia.

Aboveground Growth Measurements

At the conclusion of the experiment, samples of youngest fully expanded leaves were harvested for foliar nutrient analysis. The harvested leaves were dried at 80°C (176°F) to constant weight for subsequent tissue analysis. Each seedling was then cut off at ground level, or just above the lignotuber if one was present, and the entire seedling top dried at 80°C to constant weight. Reported total shoot dry weights include the weights of leaves harvested for nutrient analysis.

Chemical Methods for Analysis of Leaf Tissue and Organic Amendments

Samples were oven-dried, ground to a fine powder and total N was measured using a Carlo Erba NA 1500 analyser (Carlo Erba now Thermo Electric, Milano, Italy). Total P was measured by colorimetric analysis based on the reaction of soluble orthophosphate with molybdate-vanadate (Olsen and Dean 1965) in digests of finely ground samples prepared using Piper’s (1950) method, modified to omit perchloric acid, at high temperature [200–350°C (392–662°F)]. Total K was measured in the same acid digests by diluting to an appropriate range and analyzing with a Varian SpectrAA 200 atomic absorption spectrophotometer (Varian Australia, Mulgrave, VIC, Australia) in flame photometry mode (Knudsen et al. 1982).

Root Density Measurements

At the conclusion of the experiment, roots were harvested at three sample depths in each soil profile, to the east side of the transplanted root ball, using a metal core sampler [150 mm long x 25 mm (6 in x 1 in) diameter]. The first core was taken at the surface [0–150 mm (0–6 in)], the second from 180–330 mm (7–13 in), and the third from 350–500 mm (14–20 in). The core samples were washed through a series of sieves and the roots recovered. Root length was measured using a desktop scanner in conjunction with Rhizo V 4.1 software (Regent Instruments, Québec, Canada). The total root length in the samples at each depth was converted to root length density (L_V cm cm⁻³) by dividing by the core volume, which had been calculated from the dimensions of the sampler.

Statistical Analysis (Shoot and Root Growth)

The unamended sand treatment could not be included in ANO-VAs because of incomplete replication, hence the analysis of the data was done in two stages. First, an ANOVA was performed using all the fully replicated treatments. Means comparisons were made using Fisher’s least significant difference (P < 0.05). Then, to allow comparison of the unamended sand treatment with the organic-amended treatments, Dunnett’s multiple comparisons with the unamended sand as a control was used. If a significant treatment effect was found, Dunnett’s paired t-tests (P < 0.05) were then performed comparing the control with each treatment mean, adjusting for multiple comparisons. Minitab V. 12.0 statistical software (Minitab Inc., State College, PA) was used for all these analyses. Since the shoot dry weight and root density data was not normally distributed, a log_e transformation was performed and comparisons of treatment effects using ANOVA were done using this transformed data. Where transformation was used the tables of means show the back-transformed data. Since the assumptions for ANOVA were met for all foliar data sets, these data were analyzed on the original scale of measurement. There were no significant three way interactions and no significant blocking effects.