Base cation availability and leaching after nitrogen fertilisation of a eucalypt plantation.
Fertilisation of many Australian forest soils with phosphorus (P), nitrogen (N), and in some cases other nutrients, is necessary to achieve high growth rates of eucalypt plantations. (Cromer et al. 1993; Bennett et al. 1996; Cromer 1996; Judd et al. 1996; Mendham et al. 2002; Smethurst et al. 2004). However, concerns have been raised that increasing usage of nitrogenous fertilisers might have detrimental effects on base cation availability (Smethurst et al. 2004; Ringrose and Neilsen 2005a, 2005b; Stanturf and Stone 1994). A survey of concentrations of exchangeable and solution pools of cations in surface soils (04).1 m) of several N fertiliser experiments in Australian plantation forests added to this concern (Mitchell and Smethurst 2004). However, for some time after N and P fertilisation, elevated concentrations of base cations (Ca, Mg, and K) in soil solution were observed at some of these sites. This effect was particularly noticeable in the surface soil at one site, where it appeared to promote leaching of base cations to lower depths. Increased soil solution concentrations and leaching of base cations after N fertilisation is a well-recognised phenomenon internationally (Adams et al. 1997; Edwards et al. 2002), but it has received little attention in Australian soils, including those supporting eucalypt plantations, and the implications of elevated soil solution concentrations have not been considered in relation to forest nutrition.
Hence, while N fertilisation might improve the availability of cations in the short term, over the longer term it might promote leaching losses of cations. Other studies also have reported short-term increases in concentrations of cations in solution after the application of N, P, or K fertilisers, followed by increased leaching (Otehere-Boateng and Ballard 1978; Hingston and Jones 1985; Feger 1992; Khanna et al. 1992). We currently lack a reliable basis for interpreting base cation availability to forest plantations from measurements of exchangeable concentrations in soil. Instead, a means of interpreting soil solution concentrations of base cations is developing (Smethurst et al. 2007). These interpretations also needed to be complimented by knowledge of the temporal patterns of short- and long-term changes in these concentrations for soils supporting Australian eucalypt plantations.
Nitrogen fertilisation has been observed to increase base cation concentrations in water draining through soil profiles in forest ecosystems (e.g. Ring et al. 2006). However, concentrations of ions in drainage waters (usually collected by suction samplers) are usually more dilute than those surrounding roots during most periods of uptake because rapidly percolating soil water inadequately equilibrates with the soil solid phase (Litaor 1988; Smethurst 2000). The paste method, e.g. as recently applied to base cations in a Pinus radiata plantation (Smethurst et al. 2007), minimised that problem, but that study did not quantify the temporal availability of base cations as affected by N fertilisation.
Re-fertilisation with N of a fertiliser experiment in a Eucalyptus nitens plantation provided an opportunity to better define these temporal patterns of nutrient concentrations in soil solution for the subsequent year. By using a tree growth and water balance model, we also predicted maximum potential leaching fluxes of several major nutrients.
Materials and methods
Some properties and other characteristics of the Westfield fertiliser experiment are summarised in Tables 1 and 2 or previously reported (Smethurst et al. 1997, 2001, 2003; Misra et al. 1998a, 1998b; Cromer et al. 2002; Resh et al. 2003). Westfield has a clay-loam surface soil and total profile depth of approximately 0.9 m to partially weathered mudstone. The previous crop of P. radiata D. Don, which probably never received fertiliser, had replaced a high quality native eucalypt forest dominated by E. obliqua and E. regnans. The plantation was 9 years old at the start of this re-fertilisation experiment and there were 1430 stems/ha. Rooting depth was at least 0.9m (Misra et al. 1998a). An automatic weather station was installed immediately adjacent to the experiment in an open area. Sensors attached to the weather station monitored rainfall, solar radiation, and a wet and dry bulb temperature, from which relative humidity was calculated. The weather station recorded and summarised hourly and daily data. Long-term, average annual rainfall was 1384mm (Baillie and Smethurst 2002).
Fertiliser treatments were arranged in a randomised block design with 5 replicates. An unfertilised control treatment (T1) was compared to 5 fertilised treatments. Only 3 of the original 6 fertiliser treatments are reported here (T1, T2, T5). Details of the rates of N, P, and Ca applied in fertilisers at Westfield are shown in Table 3. No fertiliser was ever applied to treatment T1. Nitrogen fertiliser was broadcast applied at 2 and 26 months after planting as ammonium sulfate (20.5% N, 24.3% S) and as urea (46% N) for all subsequent fertiliser applications at 72 and 108 months. Phosphorus fertiliser was broadcast-applied to fertilised treatments in the form of triple superphosphate (20% P and 15% Ca) at 2 and 26 months only. The cumulative rates of N fertiliser applied to T2 are comparable with those currently used operationally during one rotation by one local plantation manager, but the cumulative rates used in T5 would not be attained for several rotations operationally.
Sampling and measurements
Prior to fertilisation at 108 months, soil samples were collected (October 2001) at 3 depths (0-0.1, 0.1-0.3, and 0.3-0.6m) from 4 replicates of treatments T1, T2, and T5 (0, 550, and 1600 kg N/ha pre-fertilisation) by randomly taking 20 soil cores (20mm diameter) of 0-0.1 m soil, and 10 soil cores (20mm diameter) of 0.1-0.3 and 0.3-0.6m soil. These samples were combined within a depth for each plot. In the same manner, soil samples were collected on 9 occasions during the 13-month period post-fertilisation (Table 4). Note that a cumulative N fertiliser rate of ~400 kg N/ha maximised growth to 69 months of age, with half the rate applied at both 2 and 26 months (Cromer et al. 2002), and that later applications of N did not significantly increase tree growth (Smethurst et al. 2003).
Soil solutions were extracted from a portion of the field-moist soils using the paste method (Smethurst et al. 1997). The ratio of deionised water (mL) to field moist soil (g) was 80:240 and the resulting solutions were analysed for pH, Ca, K, Ms, N[H.sub.4], and N[O.sub.4]. These nutrient concentrations were converted to those in the undiluted soil solution using an iterative solution of a mass-balance equation and measured solid-liquid phase partition coefficients ([K.sub.d], Smethurst et al. 2001). We assumed that nitrate had no interaction with the solid phase, i.e. [K.sub.d] = 0, but [K.sub.d] values for most samples were in the ranges 0.2-1.2 for N[H.sub.4], 0.4-1.0 for Ca, 6.0-7.7 for Mg, and 3.3-3.8 for K.
Estimation of drainage
Drainage D (mm) was simulated on a daily basis using the simple model:
D = R + S - C - [E.sub.t] - [E.sub.ic] - [E.sub.u]
where R (mm) is rainfall; S (mm) is the initial amount of soil water; C (mm) is the water storage capacity of the soil (95 nun for the 0.6 m soil depth); [E.sub.t] is transpiration from a dry tree canopy and of intercepted rainfall; [E.sub.ic] is evaporation from a wet tree canopy; and [E.sub.u] is evaporation directly from the forest floor. We have assumed that runoff rarely occurred because there were high rates of water uptake by trees, high infiltration rates, non-saturated soils, and low slope (4-6%). Rainfall was measured on site by an automatic weather station. The capacity of this soil to store water was estimated from values reported in the literature for soil of a similar texture as Westfield (Greacen and Williams 1983). Values of [E.sub.t], [E.sub.ic], and [E.sub.u] were predicted on a daily basis by the mechanistic growth prediction model, CABALA, that has been validated for light use and water balance in temperate eucalypt plantations (Battaglia et al. 2004). Based on crop and environmental conditions, this model simulates daily evapotranspiration, water uptake from several soil horizons, and leaching below the root-zone. Simulated tree growth was comparable to measured tree growth. Maximum potential leaching of cations (Ca, Mg, K, and N[H.sub.4]) and N[O.sub.3] moving below 0.6 m was estimated from total predicted drainage (for each month from October 2001 to October 2002) and calculated concentrations of ions in soil solution in the 0.3-0.6m depth. This estimate of leached nutrients probably represents the maximum potential, because concentrations of N[H.sub.4] and N[O.sub.3] in leachate collected in suction samplers during the first year after fertilisation were generally lower than those measured in soil solutions sampled by the paste method (Smethurst et al. 1997). For months when predicted drainage occurred and samples were not collected, nutrient concentrations from the closest sampling date were used.
Results for each sampling date were analysed for significant differences between treatment means using the analysis of variance and covariance procedures of SAS system for Windows (version 8) and Statgraphics Plus 5. Means were compared using an l.s.d, at P = 0.05, unless otherwise indicated.
Rainfall, drainage, and soil water content
Total rainfall for the 13-month period of the experiment (October 2001-October 2002) was 1610mm (Fig. 1), compared with a long-term average of 1384mm (Baillie and Smethurst 2002). Compared with the long-term data, there was a moderately wet spring and early summer (October-mid-January 2001), a dry late summer and autumn (mid-January-May 2002), and a very wet winter and early spring (June-September 2002). The first main drainage phase occurred during October and early November 2001, with little or no further drainage until June 2002. Thereafter, drainage occurred regularly until the end of the experiment in October 2002 (Fig. 1).
The trend in gravimetric soil moisture contents in the 0-0.1 m soil layer followed a similar pattern to that of total monthly rainfall (Fig. 2). They ranged from a maximum of approximately 65% during late winter-early spring (2002) to a low of approximately 40% during the autumn (2001) dry period. In the other 2 soil layers (0.1-0.3 and 0.3-0.6 m), moisture contents were relatively stable at approximately 40-45% (Fig. 2).
Solution pH before re-fertilisation was significantly lower in the T5 treatment than T1 and T2 in all depths sampled (Fig. 3). For 180 days post-fertilisation, solution pH declined further in both T2 and T5 in 0-0.1 m soil, after which solution pH increased and approximately reached pre-fertilisation levels by day 380. Similar trends were observed in the other two soil depths.
Covariance analysis of 0-0.1 m soil solutions for pH and nutrients on the day of maximum treatment effect indicated that these effects were significant in all cases, i.e. P= 0.02 for pH on day 28, P=0.04 for Ca on day 139, P=0.05 for Mg on day 139, P=0.00 for K on day 63, P=0.02 for N[H.sub.4] on day 28, and P = 0.04 for N[O.sub.4] on day 232. Hence, the trends shown in Figs 3-8 can be attributed largely to the effects of re-fertilisation rather than a cumulative effect of previous fertiliser applications.
Solution Ca, Mg, and K
Pre-fertilisation, concentrations of Ca in soil solution were significantly higher in T5 than T1 and T2 in all 3 soil layers sampled (Fig. 4). This probably reflected the application of superphosphate fertiliser (300kgP/ha, 15% Ca) to the T5 treatment during the first 2 years of the plantation. Solution Mg concentrations were similar across all treatments in the soil layers sampled pre-fertilisation (Fig. 5), but solution K concentrations in the 0-0.1 m soil layer were significantly lower in T2 than T1 and T5 (Fig. 6). There were no significant differences in soil solution K concentrations between treatments in the other soil layers sampled.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Post-fertilisation, in 0-0.1 m soil of the fertilised treatments (T2 and T5), concentrations of Ca increased compared with pre-fertilisation concentrations (significant compared with T1 after and including the sampling at 60 days). Approximately 140 days after re-fertilisation, concentrations of Ca peaked at about 4 times pre-fertilisation levels (Fig. 4). By 120 days, Ca concentrations had started to decline. By 383 days, Ca concentrations were approaching pre-fertilisation levels and those in the TI treatment, but those in T5 remained significantly higher than those in T2 and TI. Similar trends were observed for Ca concentrations in the 0.1-0.3 and 0.3-0.6 m soil depths, except that concentrations in treatment T2 were slower to increase, were highest at 320 days, and had not returned to pre-fertilisation levels by 383 days.
Post-fertilisation, soil solution concentrations of Mg followed similar trends to those of Ca. In 0-0.1 m soil of the fertilised treatments (T2 and T5), concentrations of Mg increased compared with pre-fertilisation concentrations, and the increases were significant compared with T1 at 104, 139, and 188 days. Concentrations peaked at 139 days after application at approximately 4 times higher than pre-fertilisation (Fig. 5). After 139 days, Mg concentrations declined, and by 320 days they were close to pre-fertilisation levels and those in T1. Similar trends were observed for Mg concentrations in 0.1-0.3 and 0.3-0.6 m soil.
[FIGURE 5 OMITTED]
Likewise, soil solution K concentrations in 0-0.1 m soil post-fertilisation increased in treatments T2 and T5 compared with pre-fertilisation levels, but more rapidly than for solution Ca and Mg. Solution K peaked at 63 days, but only the increase in treatment T5 compared with TI was significant (Fig. 6). After 63 days, concentrations decreased to levels similar to those in T1 by the final sampling. In 0.1-0.3 m soil, concentrations of soil solution K in T2 and T5 were generally elevated compared with T1; however, none of the differences were significant. In 0.3-0.6m soil, solution K concentrations in treatments T2 and T5 increased compared with pre-fertilisation levels, to peak at 139 days after application. Only increases in T5 were significant compared with T1. After 139 days and by 383 days, solution K concentrations declined to be similar to those in T1 and pre-fertilisation levels.
Solution ammonium and nitrate
Before re-fertilisation, soil solution ammonium (N[H.sub.4]) concentrations in the 3 soil layers sampled were similarly low across all treatments (Fig. 7). Immediately after re-fertilisation, in the 0-0.1 m soil layer, there was a large significant flush of N[H.sub.4] in T2 and T5 compared with T1 (Fig. 7). After 30 days, concentrations of solution N[H.sub.4] had decreased rapidly, and were close to pre-fertilisation levels by 383 days. However, N[H.sub.4] concentrations in the fertilised treatments remained significantly higher than in T1 up to and including the sampling at 321 days. In 0.1-0.3 m soil, soil solution N[H.sub.4] concentrations in T2 and T5 increased compared with pre-fertilisation levels. There were significant peaks compared with the T1 treatment at 63 days (T5) and 139 days (T2). After peaking, solution N[H.sub.4] concentrations declined to be similar to those in T1 and to pre-fertilisation concentrations by 383 days. In 0.3-0.6m soil, concentrations of N[H.sub.4] were generally higher in the fertilised treatments, but none of the increases were significant.
[FIGURE 6 OMITTED]
Pre-fertilisation, nitrate (N[O.sub.3]) concentrations were significantly higher in T5 than T1 and T2 in all soil layers (Fig. 8). Post-fertilisation, N[O.sub.3] concentrations in the 0-0.1 m soil layer increased in T2 and T5 compared with pre-fertilisation levels and were significant compared with the T1 treatment, peaking at 139 days (Fig. 8). Nitrate concentrations in the fertilised treatments remained relatively constant until 230 days, after which there was a sharp decrease to reach pre-fertilisation levels by 383 days. In 0.1-0.3 and 0.3-0.6 m soil, soil solution N[O.sub.3] concentrations in T2 and T5 increased compared with pre-fertilisation levels, also peaking at 139 days, except for T2 in 0.3-0.6 m soil. After 139 days, solution N[O.sub.3] concentrations generally declined but remained elevated compared with pre-fertilisation levels in the sampling at 383 days, and this difference was significant in T5 compared with T1 and T2. In 0.3-0.6 m soil of T2, concentrations of N[O.sub.3] increased steadily to a peak at 320 days, after which there was a decline in concentration. Overall, N[O.sub.3] concentrations in soil solution increased by >4 mmol/L in the 0-0.1 m depth due to fertilisation and then decreased after the soil had wet up and drainage commenced. Coincident wetting up and a decrease in N[O.sub.3] concentrations suggest that leaching of N[O.sub.3] had occurred.
[FIGURE 7 OMITTED]
Maximum potential leaching of cations and N in drainage water
In the TI treatment, K was the dominant ion in water draining below 0.6 m, with smaller losses of Ca, Mg, N[H.sub.4], and N[O.sub.3] (Table 5). Nitrogen fertiliser application greatly increased the leaching of Ca, Mg, and K, with only a small increase in N[H.sub.4] leaching. Nitrate leaching was also greatly increased by N fertiliser application, i.e. from 0.1 in T1 to 263kg N[O.sub.3]-N/ha in T5 (400 kg N/ha applied at 108 months). The increased charge associated with leached N[O.sub.3] in T5 (18.8[kmol.sub.c]/ha) was approximately 39% balanced by that associated with increased leaching of cations (7.34 [kmol.sub.c]/ha).
[FIGURE 8 OMITTED]
More than 65% of the 400kg of N applied as urea may have moved below 0.6 m depth as N[O.sub.3] in the T5 treatment. However, losses of cations were <18% of the quantity held in the exchangeable phase of the 0.1-0.3 m soil layer. Pre-fertilisation soil analyses indicated that the 0-0.1 m soil layer was affected by previous fertiliser applications (Table 2), probably by promoting both cation uptake (to meet demand for increased growth) and leaching. Maximum potential leaching of Ca, Mg, and K represented approximately 10, 17, and 11%, respectively, of the quantities held in exchangeable form in the 0-0.3 m soil layer. In terms of the exchangeable cations held in the 0-0.6 m soil profile, <9% was leached.
Results supported our expectation that N fertilisation with urea would increase concentrations of Ca, Mg, and K in soil solution, which was particularly noticeable in the 0-0.1 and 0.1-0.3 m soil depths (Figs 4-6). However, analysis of soil samples collected prior to fertilisation suggested that similar treatments during the first 6 years of this plantation had already led to decreased concentrations of exchangeable cation concentrations in the 0-0.1 m depth (Table 2) and non-significant increases in soil solution concentrations of Mg and K in the 0.1-0.3 and 0.3-0.6m depths (Figs 4-6). Together, results pre- and post-fertilisation at 9 years therefore indicate that availability (as inferred by concentrations in soil solution) increased for at least 1 year in the 0-0.1 m soil depth, and for a longer period deeper in the soil profile, despite decreased concentrations of exchangeable bases due to other applications made a few years earlier. The inference of increased base cation availability with N fertilisation was supported by consistently higher concentrations of K or Mg in leaf litter in the T5 treatment compared with T1 collected on 6 occasions between 7.3 and 7.8 years of age (C. Baillie pers. comm.; foliage was not sampled).
After many years of forest growth it is not uncommon to observe decreased concentrations of base cations in surface soil exchangeable pools (Comerford et al. 1984), and the effect can be stronger if growth was enhanced by N or P fertilisation. This effect is due to 2 main mechanisms. First, there is a net uptake of base cations during a forest rotation and redistribution to above- and below-ground biomass and litter pools (Buxbaum et al. 2005). If only wood is removed at harvest, the rate of net removal of base cations from a site is relatively low compared with total biomass removal, i.e. whole-tree plus litter layer (Judd 1996). Second, increased concentrations of nitrate and sulfate anions in soil solution after N or P fertilisation can be leached and accompanied by base cations (Otchere-Boateng and Ballard 1978; Hingston and Jones 1985; Feger 1992; Khanna et al. 1992; Ring et al. 2006). Our results were consistent with both of these effects.
Hydroponic studies have been used to interpret soil solution concentrations and base cation availability in Pinus radiata plantations (Smethurst et al. 2007), and similar criteria apply to Eucalyptus globulus (c. 180 [micro]M K and 126 [micro]M Mg required for near-maximum growth during the first 2 months, A. Knowles pers. comm.). If similar criteria apply to unfertilised E. nitens (treatment T1) in the current study, concerns are raised about possibly limiting concentrations of Mg at all depths in the current study, and a trend also towards limiting concentrations of K (Figs 5 and 6). Although Ca concentrations were also very low (Fig. 4) compared with those commonly supplied in hydroponic studies, we are unsure of the critical Ca concentrations in soil solution for these species. Fertilisation of treatments T2 and T5 increased concentrations of K and Mg to well above those required for adequate growth or E. nitens. Fertilisation at Westfield boosted tree growth by approximately 20% at 8 years of age and it was maximised by a cumulative rate of N application of 400 kg N/ha (Cromer et al. 2002). It was therefore possible that N fertilisation boosted growth not only by increasing N supply, but also by increasing the supply of base cations.
We then considered if the observed temporal patterns in concentration of nutrients in soil solution were consistent with known mechanisms of nutrient cycling. Based on simple exchange theory and the assumption of uniform rapid mixing, the large input of N[H.sub.4] resulting from urea application should have led to a very rapid increase in concentrations of base cations rather than the observed gradual increases (Figs 4-6). The reasons for the observed gradual increase in soil solution Ca, Mg, and K concentrations were unclear, although similar lags have been observed using suction samplers (Ring et al. 2006). Urea should have rapidly hydrolysed in this moist environment, but its granular nature and a diffusion-limited lag in distribution of N[H.sub.4] might have resulted in less than uniform lateral and vertical distribution throughout the soil profile. Peak concentrations in base cations in surface soil might have coincided with a peak in the uniformity of N[H.sub.4] distribution.
Solution pH also decreased with a lag. This effect was probably associated with proton generation during nitrification, increased concentrations of Al in solution, and preferential leaching of Ca, Mg, and K as the balancing charge for leached N[O.sub.3]. Acidification is usually also associated with significant leaching of other ions, i.e. H, Al, organic anions, sulphate, and chloride (Ring 2004). Elevated concentrations of aluminium in the fertilised treatments were confirmed by analysis of several paste extracts (data not shown), but other ions were not measured in this study.
After peaking, concentrations of base cations in solution in the 0-0.1 m soil layer decreased only after the soil water content had increased and substantial drainage had occurred (Figs 1, 4-8). Hence, movement of cations down the soil profile was probably the main mechanism responsible for the observed decline in concentrations after 240 days, although tree uptake would have also contributed, and possibly microbial immobilisation and precipitation of secondary minerals containing base cations.
Nitrogen fertilisation increased concentrations of N[H.sub.4] in surface soil (0-0.1 m) solutions for at least 10 months after application. Hydrolysis of the applied urea resulted in a rapid (> 100-fold) increase in soil solution N[H.sub.4] of 1.71 mmol/L within 30 days after fertiliser application compared with the initial concentrations in the fertilised T5 treatment (Fig. 7). Similarly, Otchere-Boateng and Ballard (1978) reported soil solution N[H.sub.4] of 1.9mmol/L after applying 448 kg N/ha as urea to soils under second-growth forest of predominantly Douglas fir (Pseudotsuga menziesii) of varying ages. Also, Smethurst et al. (2001) reported increases in soil solution N[H.sub.4] of 5.5 mmol/L at the same site after earlier applications (2 and 26 months after planting) using 600 kg N/ha as ammonium sulfate. Because N turnover can be rapid, between application and the first sampling (28 days) some fertiliser N was probably immobilised, mineralised, nitrified, and leached. Hence, the peak in N[H.sub.4] concentrations at 28 days was a measure of the net effect of these processes.
Ammonium concentrations remained relatively constant to 104 days after fertiliser application, after which there was a rapid decline to concentrations close to pre-fertilisation levels. Apart from some uptake by the trees, a considerable proportion of the N[H.sub.4] was nitrified (Fig. 8). The fate of the remaining N[H.sub.4] removed from the 0-0.1 m soil layer was unclear, but there was evidence of some movement of N[H.sub.4] to lower soil layers (0.1-0.3 m) (Fig. 7) and some was probably immobilised in microbial biomass soon after application. Also, N[H.sub.4] might have continued to exchange with Ca, Mg, and K held on poorly accessible exchange sites, thereby slowly removing N[H.sub.4] from solution and resulting in the gradual increase in the soil solution concentrations of Ca, Mg, and K in the 0-0.1 m soil layer (Figs 4-6).
How significant then was acidification and leaching of base cations induced by N-fertiliser application? Changes in solution N[O.sub.3] concentrations in the fertilised treatments are parallel to changes in soil solution pH and concentrations of Ca, Mg, and K (Figs 3-6 and 8). Significant relationships were found between concentrations of N[O.sub.3] in solution and concentrations of Ca, Mg, and K and solution pH (Table 6) as others have found (Ring et al. 2006). Continued nitrification of N[H.sub.4] could also explain why soil solution concentrations of Ca, Mg, and K increased and pH decreased for several months after fertiliser application, although some of the decline in pH was probably due to seasonal fluctuations, as solution pH also declined in the T1 treatment (Fig. 3). Ca and Mg concentrations appeared to be strongly affected by nitrification (Figs 4 and 5), as also observed by Otehere-Boateng and Ballard (1978). The release of H ions into solution with the conversion of N[H.sub.4] to N[O.sub.3] is partially responsible for observed decreases in soil pH and displacement of Ca, Mg, and K from exchange sites into soil solution (Edwards et al. 2002). A simple proton budget for our site (pH < 5) indicates that the net number of protons produced in soil per molecule of urea applied would have been -1 with N[H.sub.4] uptake, 0 with N[O.sub.3] uptake, or 1 with N[O.sub.3] leaching (Binkley and Richter 1987). This budget and our data indicate that urea application followed by N[O.sub.3] leaching was a net acidifying process.
Leaching losses of Ca and K due to fertilisation (Table 5) are likely to be small compared with the quantities that could be removed in harvested logs and subsequent site preparation. For example, nutrient budgets developed for a typical 20-year rotation of eucalypt plantation and debarked on-site, with a low intensity regeneration bum, suggest that as much as 241 kg Ca/ha and 164 kg K/ha can be lost in harvested and volatilised material (Judd 1996), compared with 90 kg Ca/ha and 30 kg K/ha leached below 0.6 m depth (Table 5). However, the converse was estimated for Mg, i.e. 8 kg Mg/ha removed in harvested wood (Judd 1996) compared to 27 kg Mg/ha leached (Table 5). Based on measurements by Adams and Attiwill (1991), annual inputs in rainfall were probably ~11 kg Ca/ha, 2.2 kg Mg/ha, 4.5kg K/ha, 0.5 kg N[H.sub.4]/ha, and 1.6 kg N[O.sub.3]/ha. Hence, in comparison with values in Table 5, rainfall inputs over a typical pulpwood rotation of 15 years could balance leaching fluxes of Ca, Mg, and K after a single application of urea at a high rate, but N leaching would not be balanced by rainfall inputs. Although not quantified, weathering inputs of Ca, Mg, and K into soil solution were also likely, but not of N. Because low N availability was likely to continue to limit wood production at this site, periodic applications of N fertiliser will be needed. Conversely, it was unclear if and when applications of Ca, Mg, or K will be needed.
The fate of nutrients leached below 0.6 m depth remains uncertain. Base cations retained in subsoils can be an important source of these nutrients during the life of a forest (Buxbaum et al. 2005; Van Rees and Comerford 1986; Berger et al. 2006), and eucalypts forests can be very deep-rooted. However, we did not assess the importance of subsoil base cation uptake at this site. We conclude that N[H.sub.4] sources of N fertilisers (including urea) potentially increase base cation availability for more than one rotation of a eucalypt plantation, which can be as short as 10 years in Australia, and even 5 years in some tropical countries. However, much of applied N will be nitrified and some leached with accompanying base cations. Management and modelling of base cation availability associated with N fertilisation will therefore need to consider both potential increases in soil solution concentrations throughout the profile for at least the current rotation and probably longer, as well as the longer term fate of leached cations, i.e. potential uptake from deeper horizons, and leaching loses beyond the root-zone.
we thank Norske-Skog Ltd for access to the experimental sites, and Craig Baillie for assistance with sample collection. We thank Ann Wilkinson for assistance with the analysis of soil solutions, Michael Battaglia for assistance with tree growth and water balance modelling, and Peter Clinton, Richard Doyle, and an anonymous reviewer for comments on an earlier version of the manuscript.
Manuscript received 7 January 2008, accepted 23 June 2008
Adams MA, Attiwill PM (1991) Nutrient balance in forests of northern Tasmania. I. Atmospheric inputs and within-stand cycles. Forest Ecology and Management 44, 93-113. doi: 10.1016/0378-1127(91) 90001-C
Adams MB, Angradi TT, Kochenderfer .IN (1997) Stream water and soil solution responses to 5 years of nitrogen and sulfur additions at the Fernow Experimental Forest, West Virginia. Forest Ecology and Management 95, 79-91. doi: 10.1016/S0378-1127(96)03990-4
Baillie CC, Smethurst PJ (2002) Climate in the Florentine Valley. Cooperative Research Centre for Sustainable Production Forestry, Technical Report 72, Hobart, Tas., Australia.
Battaglia M, Sands P, White D, Mummery D (2004) Cabala: a linked carbon, water and nitrogen model of forest growth for silvicultural decision support. Forest Ecology and Management 193, 251-282. doi: 10.1016/j. foreco.2004.01.033
Bennett LT, Weston C J, Judd TS, Attiwill PM, Whiteman PH (1996) The effects of fertilizers on early growth and foliar nutrient concentrations of three plantation eucalypts on high quality sites in Gippsland, southeastern Australia. Forest Ecology and Management 89, 213-226. doi: 10.1016/S0378-1127(96)03848-0
Berger TW, Swoboda S, Pruhaska T, Glatzel G (2006) The rule of calcium uptake from deep soils for spruce (Picea abies) and beech (Fagus sylvatica). Forest Ecology and Management 229, 234-246. doi: 10.1016/j.foreco.2006.04.004
Binkley D, Richter D (1987) Nutrient cycles and H+ budgets of forest ecosystems. Advances in Ecological Research 16, 1-51. doi: 10.1016/ S0065-2504(08)60086-0
Blakemore LC, Searle PL, Daly BK (1987) Methods for chemical analysis of soils. New Zealand Soil Bureau, Scientific Report 80, Lower Hutt, New Zealand.
Buxbaum CA, Nowak CA, White EH (2005) Deep subsoil nutrient uptake in potassium-deficient, aggrading Pinus resinosa plantation. Canadian Journal of Forest Research 35, 1978-1983. doi: 10.1139/x05-102
Comerford NB, Kidder G, Mollitor AV (1984) Importance of subsoil fertility to forest and non-forest plant nutrition. In 'Forest Soils and Treatment Impacts: Proceedings of Sixth North American Forest Soils Conference'. (Ed. EL Stone) pp. 381-404. (University of Tennessee: Knoxville, TN)
Cromer RN (1996) Silviculture of eucalypt plantations in Australia. In 'Nutrition of eucalypts'. (Eds PM Attiwill, MA Adams) pp. 259-273. (CSIRO Publishing: Collingwood, Vic.)
Cromer RN, Cameron DM, Rance S J, Ryan PA, Brown M (1993) Response to nutrient in Eucalyptus grandis. 2. Nitrogen accumulation. Forest Ecology and Management 62, 231-243. doi: 10. 1016/0378-1127(93) 90052-O
Cromer RN, Tumbull CRA, LaSala AV, Smethurst P J, Mitchell AD (2002) Eucalyptus growth in relation to combined nitrogen and phosphorus fertiliser and soil chemistry in Tasmania. Australian Forestry 65, 87-95.
Edwards PJ, Wood F, Koehenderfer JN (2002) Baseflow and peakflow chemical responses to experimental applications of ammonium sulphate to forested watersheds in north-central West Virginia, USA. Hydrological Processes 16, 2287-2310. doi: 10.1002/hyp.1000
Feger KH (1992) Nitrogen cycling in two Norway spruce (Picea abies) ecosystems and effects of a [(N[H.sub.4]).sub.2]S[O.sub.4] addition. Water, Air, and Soil Pollution 61, 295-307. doi: 10.1007/BF00482612
Greacen EL, Williams J (1983) Physical properties and water relations. In 'Soils: an Australian viewpoint', pp. 499-530. (CSIRO: East Melbourne, Vic.)
Hingston FJ, Jones MS (1985) Changes in the composition of soil solutions resulting from application of fertilizer to jarrah forest in south-western Australia. Australian Forest Research 15, 293-308.
Isbell RF (1996) 'The Australian soil classification.' (CSIRO Publishing: Collingwood, Vic.)
Judd TS (1996) Simulated nutrient losses due to timber harvesting in highly productive eucalypt forests and plantations. In 'Nutrition of eucalypts'. (Eds PM Attiwill, MA Adams) pp. 249-258. (CSIRO Publishing: Collingwood, Vic.)
Judd TS, Bennett LT, Weston CJ, Attiwill PM, Whiteman PH (1996) The response of growth and foliar nutrients to fertilizers in young Eucalyptus globulus (Labill.) plantations in Gippsland, southeastern Australia. Forest Ecology and Management 82, 87-101. doi: 10.1016/0378-1127(95)03693-8
Khanna PK, Raison RJ, Falkiner RA, Willett IR, Connell MJ (1992) Effect of NPK fertilisation on the chemistry of a yellow podsolic soil under Pinus radiata. Forest Ecology and Management 52, 65-85. doi: 10.1016/0378-1127(92)90496-V
Litaor MI (1988) Review of soil solution samplers. Water Resources Research 24, 727-733. doi: 10.1029/WR024i005p00727
Mendham DS, Smethurst PJ, Holz GK, Menary RC, Grove TS, Weston C, Baker T (2002) Soil analyses as indicators of P status in young Eucalyptus nitens and E. globulus plantations. Soil Science Society of America Journal 66, 959-968.
Misra RK, Turnbull CRA, Cromer RN, Gibbons AK, LaSala AV (1998a) Below and above ground growth of Eucalyptus nitens in a young plantation I. Biomass. Forest Ecology and Management 106, 283-293. doi: 10.1016/S0378-1127(97)00339-3
Misra RK, Turnbull CRA, Cromer RN, Gibbons AK, LaSala AV, Ballard LM (1998b) Below and above ground growth of Eucalyptus nitens in a young plantation II. Nitrogen and phosphorus. Forest Ecology and Management 106, 295-306. doi: 10.1016/S0378-1127(97)00320-4
Mitchell AD, Smethurst PJ (2004) Surface soil changes in base cation concentrations in fertilized hardwood and softwood plantations in Australia. Forest Ecology and Management 191, 253-265. doi: 10.1016/j.foreco.2003.12.011
Otchere-Boateng J, Ballard TM (1978) Urea fertilizer effects on dissolved nutrient concentrations in some forest soils. Soil Science Society of America Journal 42, 503-508.
Resh SC, Battaglia M, Worledge D, Ladiges S (2003) Coarse root biomass for eucalypt plantations in Tasmania, Australia: sources of variation and methods for assessment. Trees 17, 389-399.
Ring E (2004) Experimental N fertilization of scots pine: effects on soil-solution chemistry 8 years after final felling. Forest Ecology and Management 188, 91-99. doi: 10.1016/j.foreco.2003.07.020
Ring E, Jacobson S, Nohrstedt H (2006) Soil-solution chemistry in a coniferous stand after adding wood ash and nitrogen. Canadian Journal of Forest Research 36, 153-163. doi: 10.1139/x05-242
Ringrose C, Neilsen WA (2005a) Growth response of Eucalyptus regnans and soil changes following periodic fertilization. Soil Science Society of America Journal 69, 1806-1812. doi: 10.2136/sssaj2004.0237
Ringrose C, Neilsen WA (2005b) Growth responses of Pinus radiata and soil changes following periodic fertilization. Soil Science Society of America Journal 69, 1799-1805. doi: 10.2136/sssaj2003.034l
Smethurst PJ (2000) Soil solution and other soil analyses as indicators of nutrient supply: a review. Forest Ecology and Management 138, 397-411. doi: 10.1016/S0378-1127(00)00426-6
Smethurst PJ, Baillie C, Cherry M, Holz G (2003) Fertilizer effects on LAI and growth of four Eucalyptus nitens plantation. Forest Ecology and Management 176, 531-542. doi: 10.1016/S0378-1127(02)00226-8
Smethurst PJ, Herbert AM, Ballard LM (1997) A paste method for estimating concentrations of ammonium, nitrate and phosphate in soil solution. Australian Journal of Soil Research 35, 209-225. doi: 10.1071/ S95040
Smethurst PJ, Herbert AM, Ballard LM (2001) Fertilization effects on soil solution chemistry in three eucalypt plantations. Soil Science Society of America Journal 65, 795-804.
Smethurst P, Holz H, Moroni M, Baillie C (2004) Nitrogen management in Eucalyptus nitens plantations. Forest Ecology and Management 193, 63-80. doi: 10.1016/j.foreco.2004.01.023
Smethurst PJ, Knowles A, Wilkinson A, Churchill K, Lyons A (2007) Soil and foliar chemistry associated with potassium deficiency in Pinus radiata. Canadian Journal of Forest Research 37, 1093-1105. doi: 10.1139/X06-307
Stanturf JA, Stone EL (1994) Loss of nitrogen and bases after fertilization of second-growth hardwood forest soils. Forest Ecology and Management 65, 265-277. doi: 10.1016/0378-1127(94)90175-9
Van Rees KCJ, Comerford NB (1986) Vertical root distribution and strontium uptake of a slash pine stand on a Florida Spodosol. Soil Science Society of America Journal 50, 1042-1046.
A.D. Mitchell (A,C) and P.J. Smethurst (A,B,D)
(A) CRC Forestry Ltd, Private Bag 12, Hobart, Tas. 7001, Australia.
(B) CSIRO Forest Biosciences, Private Bag 12, Hobart, Tas. 7001, Australia.
(C) Current address: NZ Sport Turf Institute (Inc), PO Box 347, Palmerston North, New Zealand.
(D) Corresponding author. Email: Philip.Smethurst@csiro.au
Table 1. Some characteristics of the Westfield fertiliser experiment Soil surface texture Clay-loam Parent material Mudstone Australian Soil Classification (A) Kurosol Latitude 42[degrees]39'S Longitude 146[degrees]28' Elevation (m a.s.l.) 430 Average rainfall (mm/year) 1370 Current crop Eucalyptus nitens Year planted 1992 Stocking (stems/ha) 1430 (A) Isbell (1996). Table 2. Some chemical characteristics of 3 treatments (T1, T2, and T5) of the Westfield fertiliser experiment prior to re-application of fertiliser at age 9 years Within rows, means followed by the same letter are not significantly different (l.s.d. at P = 0.05). Refer to Table 3 for treatment definitions Characteristic Depth (m) T1 T2 T5 Soil pH 0-0.1 4.53a 4.50a 4.22a EC (dS/m) 0-0.1 0.07ab 0.06a 0.08b ECEC (A) ([cmol.sub.c]/kg) 0-0.1 14.73a 13.50a 14.36a Exch. Ca ([cmol.sub.c]/kg) 0-0.1 2.84a 2.22a 2.15a 0.1-0.3 1.33a 1.59a 1.13a 0.3-0.6 0.79a 0.78a 0.82a Exch. Mg ([cmol.sub.c]/kg) 0-0.1 1.24a 0.69b 0.496 0.1-0.3 0.60a 0.55a 0.34a 0.3-0.6 0.31a 0.28a 0.31a Exch. K ([cmol.sub.c]//kg) 0-0.1 0.38a 0.27b 0.266 0.1-0.3 0.26a 0.20a 0.20a 0.3-0.6 0.21a 0.21a 0.25a KC1-extractable (mmol/kg): N[H.sub.4] 0-0.1 0.24a 0.28a 0.576 N[O.sub.3] 0-0.1 >0.01a 0.02a 0.486 Blakemore et al. (1987). (^) Exchangeable cations (Ca, Mg, K, Na) plus exchangeable acidity (H, A1) Blakemore et al. (1987). Table 3. Fertiliser history (N : P : Ca; kg/ha elemental) of the treatments used in this study Treatment code 2 months 26 months 72 months T1 0 : 0 : 0 0 : 0 : 0 0 : 0 : 0 T2 75 : 37.5 : 28 75 : 37.5 : 28 400 : 0 : 0 T5 600 : 300 : 225 600 : 300 : 225 400 : 0 : 0 Treatment code 108 months Total T1 0 : 0 : 0 0 : 0 : 0 T2 400 : 0 : 0 950 : 75 : 0 T5 400 : 0 : 0 2000 : 600 : 450 Table 4. Schedule of sampling dates and depths Date Days after re-fertilisation Depths sampled (m) Oct. 2001 Pre-fertilisation 0-0.1, 0.1-0.3, 0.3-0.6 Nov. 2001 28 0-0.1 Dec. 2001 63 0-0.1, 0.1-0.3, 0.3-0.6 Jan. 2002 104 0-0.1 Feb. 2002 139 0-0.1, 0.1-0.3, 0.3-0.6 April 2002 188 0-0.1 May 2002 232 0-0.1 July 2002 279 0-0.1, 0.1-0.3, 0.3-0.6 Aug. 2002 321 0-0.1, 0.1-0.3, 0.3-0.6 Oct. 2002 383 0-0.1, 0.1-0.3, 0.3-0.6 Table 5. Estimated leaching losses (kg/ha) of Ca, Mg, K, N[H.sub.4]-N, and N[O.sub.3]-N below 0.6 m in the T1 and T5 treatments for months when drainage occurred Month Ca Mg K T1 T5 T1 T5 T1 T5 Oct. 0.2 6.9 0.2 3.1 1.5 3.70 Nov. 0.1 2.5 <0.1 0.6 0.3 1.1 Jan. <0.1 0.3 <0.1 0.1 <0.1 0.1 June 0.7 18.9 0.3 3.2 1.8 5.7 July 0.8 20.9 0.4 5.7 2.0 6.3 Aug. -- 12.1 0.2 3.3 0.8 2.8 Sep. 0.3 19.4 0.3 5.7 2.0 7.0 Oct. 0.1 9.4 0.2 2.8 1.0 3.4 Total 2.2 90.4 1.6 26.5 9.3 30.1 Month N[H.sub.4]-N N[O.sub.3]-N T1 T5 T1 T5 Oct. <0.1 -0.1 -- 18.1 Nov. -- 0.3 -- 7.7 Jan. -- <0.1 -- 1.0 June <0.1 0.3 <0.1 57.6 July <0.1 0.3 <0.1 63.7 Aug. <0.1 0.2 -0.1 34.9 Sep. <0.1 0.5 -- 53.9 Oct. <0.1 0.3 -- 26.2 Total 0.1 2.0 0.1 263.0 Table 6. Results of linear regression analysis of the relationship between soil solution concentrations of N[O.sub.3] and pH or concentrations of Ca, Mg, and K in the fertilised treatments All relationships were significant at P<0.001 Depth Variable Coefficient [r.sup.2] (m) correlation (r) 0-0.1 pH -0.75 0.56 Ca 0.95 0.91 Mg 0.89 0.79 K 0.79 0.62 0.1-0.3 pH -0.65 0.42 Ca 0.94 0.88 Mg 0.83 0.69 K 0.79 0.61 0.3-0.6 pH -0.76 0.57 Ca 0.90 0.81 Mg 0.88 0.77 K 0.65 0.43
|Printer friendly Cite/link Email Feedback|
|Author:||Mitchell, A.D.; Smethurst, P.J.|
|Publication:||Australian Journal of Soil Research|
|Article Type:||Technical report|
|Date:||Aug 1, 2008|
|Previous Article:||Using poultry litter biochars as soil amendments.|
|Next Article:||Effects of different soil management practices on winter wheat yield and N losses on a dryland loess soil in China.|