Plant and Soil

, Volume 346, Issue 1, pp 63–78

Effectiveness of wood ash containing charcoal as a fertilizer for a forest plantation in a temperate region

Authors

  • Martín Santalla
    • Sustainable Forest Management UnitEscuela Politécnica Superior, University of Santiago de Compostela
  • Beatriz Omil
    • Sustainable Forest Management UnitEscuela Politécnica Superior, University of Santiago de Compostela
  • Roque Rodríguez-Soalleiro
    • Sustainable Forest Management UnitEscuela Politécnica Superior, University of Santiago de Compostela
    • Department of Crop ProductionEscuela Politécnica Superior, University of Santiago de Compostela
    • Sustainable Forest Management UnitEscuela Politécnica Superior, University of Santiago de Compostela
    • Department of Soil Science and Agricultural ChemistryEscuela Politécnica Superior, University of Santiago de Compostela
Regular Article

DOI: 10.1007/s11104-011-0794-y

Cite this article as:
Santalla, M., Omil, B., Rodríguez-Soalleiro, R. et al. Plant Soil (2011) 346: 63. doi:10.1007/s11104-011-0794-y

Abstract

Amendment of forest soils with mixed wood ash (MWA) generated in biomass power plants can prevent the depletion of soil nutrients that results from the intensive harvesting of forest plantations. Unlike fly wood ash, MWA contains charcoal and is characterized by a lower release of nutrients, so that it might be useful as a long term source of nutrients and soil organic matter. However, in order to use MWA as a fertilizer in forest systems, its effectiveness as regards supplying P and N must be improved. These aspects were studied in a 4 year-trial carried out in a Pinus radiata plantation. MWA was added alone or with mineral P, and the results were compared with those obtained with a combination of Ca(OH)2 and mineral P. The application of MWA together with mineral P fertilizer increased the nutrient supply to the trees, as revealed by the changes in nutrient concentrations, lower values of resorption efficiencies and improved tree growth. The results showed that the amounts of Ca, Mg and K supplied by the MWA were suitable for maintenance of soil reserves. However, the presence of charcoal may have decreased the availability of P. The application of the MWA led to lower soil N mineralization rates and mineral N concentrations, which may affect N-limited systems. The use of density-dependent single tree increment models enabled the positive effects on tree growth of fertilization and thinning to be distinguished. For the treatments supplemented with mineral P, multiplicative factors of 1.13 to 1.15 can be applied to obtain post-thinning predictions of 4-year single-tree basal area increments. Although MWA can be used as a long term source of nutrients, charcoal temporarily reduces the availability of P and N.

Keywords

Wood ashCharcoalFertilizationPhosphorusNitrogenDecompositionLitterfallNutrient cyclingPinus radiataTree growth

Introduction

Intensification of forest management in specific areas is necessary to ensure sufficient wood and energy supplies to meet the current demand. However, in some areas intensive forest management leads to soil degradation due to the depletion of nutrients and soil organic matter. This is the case in intensively managed forests in temperate and tropical forests where application of fertilizers is therefore necessary to replenish the nutrients and prevent forest decline from occurring.

Different studies have shown that the use of wood ash (derived from biomass combustion in power plants) as a fertilizer is a suitable strategy for improving nutritional status and forest growth (e.g. Augusto et al. 2008), and also contributes to integrated environmental management involving the forestry, industry and energy sectors. Two types of wood ash are produced in power plants—fly ash and bottom ash—and are distinguished by their reactivity and heavy metal contents. Fly powdered ash, typically light-grey, is captured from boiler emissions, in cyclone separators, and contains large amounts of nutrients (except N), which are present as dissolvable salts and are rapidly released. In contrast, bottom or boiler ash, usually referred to as mixed wood ash (MWA), and which is produced in wood fired furnaces, is less reactive and contains lower amounts of heavy metals than fly ash because many of the elements are present in partially burnt organic compounds and charcoal (in some case more than 40%).

Fly wood ash contains large amounts of readily available elements (K, Ca and Mg), and its application to forest soils has been widely reported to be beneficial for tree nutrition, mainly in boreal forests (e.g. Helmisaari et al. 2009) and to a much lesser extent in temperate regions (e.g. Solla-Gullón et al. 2006). As a result of the lower release of elements, the application of MWA may prevent rapid, undesirable changes in soil forest properties, maintain the availability of these elements for longer (Solla-Gullón et al. 2008) and also reduce the risk of heavy metal contamination (Omil et al. 2007; Reimann et al. 2008). In addition, charcoal has been found to have beneficial effects on plant growth (i.e. Atkinson et al. 2010), owing to improvements in physical (water retention and aggregation), chemical (increased soil pH and CEC) and biological (mycorrhizal associations, higher microbial activity) properties of soil. Moreover, charcoal is relatively recalcitrant and can therefore act as a long-term sink for atmospheric CO2 (e.g. Krull et al. 2006).

Different studies on forest soils treated with wood ash (e.g. Augusto et al. 2008) have shown moderate responses in soil available and plant P, possibly owing to the low solubility of ash compounds containing P (Schiemenz and Eichler-Lobermann 2010) and to the strong influence of pH on P biochemistry. This aspect is particularly important in acid soils in temperate forests, in which P is the most limiting nutrient. The combined application of fly ash and mineral P fertilizer, which has been shown to overcome this limitation in agricultural land (Merino et al. 2006; Ferreiro et al. 2011), has not yet been tested in P-limited forests. However, the dynamics of P after the application of wood ash containing charcoal are not easily predicted. On one hand, some studies have shown that wood ash can enhance the uptake of P, because the P is solubilized by ectomycorrhizal fungi that colonise wood ash (Warnock et al. 2007; Makoto et al. 2010), and also because of changes in root morphology (Majdi et al. 2008). On the other hand, charcoal can absorb inorganic and organic P in its pores and may favour the formation of poorly soluble phosphates, which would subsequently be released (Laird et al. 2010).

Although wood ash contains low amounts of available N, it can exert a considerable influence on the N cycle, favouring N mineralization in soils rich in organic matter (Saarsalmi et al. 2010). The application of ash containing charcoal has the potential to modify the N dynamics through different mechanisms, some working in opposite directions (Clough and Condron 2010). The ash generated in wildfires can enhance nitrification as a result of the increased pH and, in certain environments, as a result of charcoal-induced adsorption of phenols and terpenes, which inhibit nitrifying bacteria (Ball et al. 2010). On the other hand, the high C:N ratio in the charcoal may reduce N mineralization, with the subsequent effects on plant N uptake (e.g. Steiner et al. 2008).

In addition to these effects, wood ash has been shown to have an indirect influence on nutritional status, through changes in soil biological activity. Thus, the increase in pH after application of fly wood ash may result in increased decomposition of the soil organic matter (Perkiömäki et al. 2004). The application of MWA may alter the soil microbial activity in different ways because of the high content of charcoal in the ash. Charcoal contains highly recalcitrant C molecules that may provide a poorly decomposable C substrate, thus leading to decreased C mineralization of native soil C, and deceleration of C cycling in the soil (Liang et al. 2010). However, this is not clear because charcoal particles may promote litter decomposition as a result of higher microbial activity (Wardle et al. 2008).

Furthermore, studies evaluating the response of forests to the application of wood ash are based on certain indicators, such as soil and tissue analysis and tree growth, which usually show moderate responses (e.g. Helmisaari et al. 2009). Foliar analysis does not usually reveal changes in element concentrations, which restricts interpretation of the tree response. None of the studies on forest wood ash fertilization carried out to date have evaluated the changes in litterfall in the stands treated with wood ash. Thus, litterfall monitoring may provide information about the early response of trees in terms of biomass production and internal fluxes of nutrients (Campo et al. 2007).

The silvicultural basis of fertilization after thinning is to apply readily available nutrients at a time when other competitive stresses have been reduced by the thinning, to produce growth of the remaining trees beyond the thinning response (Turner et al. 1992). The effects on growth and wood density of fertilization and thinning applied at older ages have been observed in different studies, which have shown that these treatments can dramatically increase wood volume over a short period towards the end of the rotation (Turner et al. 1992; Snowdon and Waring 1990; Nyakuengama et al. 2002). The maximum response was expected after 4 years (Snowdon and Waring 1990). Application of P significantly lowered the timber density in comparison with the control in the post-treatment period, and the application of fertilizer also increased the proportion of wood formed in mid-rotation, relative to juvenile wood (Nyakuengama et al. 2002).

There are still considerable gaps in knowledge as regards understanding the mechanisms by which fertilization of forest stands with MWA, a complex mixture of fly ash and charcoal, can influence nutrient biogeochemistry and how it affects nutrient supply to the plants. In the present study, the main advantages and limitations of MWA as a forest fertilizer were assessed in a Pinus radiata plantation, in which particular emphasis was placed on the P and N dynamics. The early response of the increased nutrient supply was evaluated by monitoring litterfall production and nutrient concentrations. Models able to distinguish the effect of the previous thinning from that of fertilization on tree growth were used.

Materials and methods

Site characteristics

The study site is located in Guitiriz (Lugo, NW Spain; UTM Lat 4.783.191, Long 598.914), at an altitude of approximately 450 m. The climate of the area is sub-humid Mediterranean with a centro-European trend. The 20 year annual average rainfall of the area is 1134 mm, and the temperature, 15.1 ºC. The wettest month is November, with an average rainfall of (157) mm, and the driest August, with (17.5) mm. The lowest mean monthly temperature (5.4 ºC) occurs in January.

The soil is developed from shale and was classified as Alumi-humic Umbrisol (IUSS Working Group WRB 2006). The soil has a fine texture and shows relatively low hydraulic conductivity. The A horizon is rich in organic matter, strongly acidic, with low CEC and low levels of available P, Ca, Mg and K (Table 1). The soil humidity and temperature regimes are Udic (mean period with partial drought, 2 months) and Mesic (mean frost-free period, 10 months), respectively.
Table 1

Selected general properties of the soil under study

Horizon

pHKCl

C (%)

C/N

CECe (cmoc kg−1)

Base Cat. Saturation (%)

Sand (%)

Clay (%)

Texture

A

3.64

4.4

18.1

10.7

14.8

40.4

17.9

Loam

Bw

3.81

2.7

20.9

13.9

17

41.8

17

Loam

The study was carried out in an area of 14 ha, in a 25-yr-old second rotation Pinus radiata plantation covering 246 ha. The site index (the total height to which dominant trees of a given species grow on a given site at 20 yr) in the plantation was 17.7 m. This is a medium-low value for the range (13 to 25 m) found in the region (Diéguez-Aranga et al. 2005). The initial foliar analyses revealed deficient concentrations of Ca and Mg, and low concentrations of P, which is common in Pinus radiata plantations in the region (Sánchez-Rodríguez et al. 2002).

Respacing was carried out in 1990, commercial thinning in July 2003, and pruning (to 5 m) in February 2004. Before treatment, the stand density was 420 trees ha−1. The plantations included young individual specimens of Betula alba, Castanea sativa and Quercus robur. The underground vegetation mainly consisted of Rubus sp., with Agrostis sp, Avenula sp. Holcus sp. Pteridium aquilinum, Ulex europaeus, Ulex gallii alyssoides, Erica cinerea, Calluna vulgaris, Cytisus sp. and Pseudorrhenatherum longifolium, all of which are common understorey species in acidic Atlantic pine plantations.

Experimental design

The experiment was based on a randomised block design with four replicates. Each block included the following four treatments:
  • C (Control): untreated, with no ash or nutrients added;

  • CaP: 493 kg of Ca ha−1 (630 kg of Ca(OH)2) + 65 kg P ha−1 (333 kg of superphosphate)

  • MWA: 7.5. Mg MWA ha−1

  • MWAP: 7.5. Mg MWA ha−1 + 65 kg P ha−1 (333 kg of superphosphate)

The plots measured 35 × 35 m, and were arranged in 4 blocks each of 4 plots (n = 16). All trees were labelled with a number.

The ash used in this study was obtained from a thermal power plant adjoining a chipboard factory (Financiera Maderera, S.A., Rábade, Spain). The ash was mainly derived from the combustion of Pinus radiata bark. The ash, which was not conditioned, was a mixture of bottom ash (rich in charcoal and other unburned organic matter) and cyclone fly ash derived from a grate fired boiler (powdered ash). The MWA was not previously sieved and contained a large amount of coarse unburned fragments. The particle size was classified as follows: <2 mm: 30%, 2–4 mm: 27.2%, 4–8 mm: 20% and >8 mm: 22.8%.

The composition of the wood-bark ash applied is shown in Table 2. The material was alkaline (pH 10.4), and the equivalent neutralizing value was 20%, expressed as percentage of standard limestone. The concentrations of trace metals were below the current limits established by the EU (European Communities 1986) for biosolids applied to agricultural soils. Analysis of the Mehlich 3 extract, considered as an indicator of element availability, revealed low contents of available trace elements. The concentrations of P, Mg and K in the MWA were 3 times lower, and the concentration of Ca was 4 times lower than in the fly powered wood ash used in previous studies (Solla-Gullón et al. 2006).
Table 2

Element composition of the wood ash used in the study and estimated amounts of each element applied at each of the doses used. A comparison with maximum loading rates for land application of sewage sludge is also included

 

Whole sample

Fine fraction from cyclone separator

Total1 (g kg−1)

Extractable2 (g kg−1)

7.5 Mg wood ash (kg)

Annual maximum loading rates for sludge-EU3 (kg ha−1)

Total1 (g kg−1)

Extractable2 (g kg−1)

pH (KCl)

11.9

     

Organic matter

561

 

4207

 

461

 

N

1.2

 

9

 

1.9

 

C/N

467.5

     

S

0.32

   

0.3

 

P

2

0.31

15

 

5.2

0.12

Ca

25.1

13.9

186

 

99.7

37.4

Mg

8.07

2.02

60

 

20.2

6.7

Na

2.4

1.2

    

K

9.8

4.41

73

 

30.2

14.5

Al

n.d.

n.d.

  

15.1

3.1

Fe

13.5

3.69

100

   

Mn

1.7

0.41

13

 

6.1

2.7

B

0.057

0.002

  

0.142

0.0017

Cd

0.00065

0.00040

0.005

39

0.004

0.002

Cr

0.083

0.013

0.62

3

0.023

0.005

Cu

0.147

0.020

1.1

12

0.073

0.024

Hg

  

0.1

 

Ni

0.054

0.008

0.40

3

0.117

0.067

Pb

0.107

0.014

0.80

15

0.017

0.008

Zn

0.241

0.044

1.8

30

0.335

0.108

1Digested with HNO3 in a microwave oven (with the exception of C, which was determined by calcination

2Extraction with Mehlich 3 reagent

3European Communities (1986)

In order to select the dose of MWA, the results of previous studies, in which the response of young plantations and the risk of heavy metal contamination were evaluated (Solla-Gullón et al. 2008; Omil et al. 2007), were taken into account. A low amount of MWA was added (7.5 Mg ha−1) because the main objective was to help counteract the export of nutrients due to conventional harvesting. This minimized the potential environmental risks associated with the application of wood ash. The main aim of adding the dose chosen was therefore to compensate the exports of the P, Ca and Mg, which are the main limiting nutrients for these plantations (Merino et al. 2005). Since the amount of P added with ash (15 kg ha−1) was not sufficient to compensate the exports during the harvesting, the MWA was supplemented with mineral P fertilizer.

The treatments were carried out in March 2004. A four-wheel-drive farm tractor (88 kW) was used to pull the spreader to apply the MWA in the thinned plantations.

Sampling and analysis of soil and plant

The study was carried out over 4 years, between March 2004 and December 2008. Soil samples, and green needles were collected annually, in October or November of each year, whereas litterfall was collected seasonally. Mineral N was determined quarterly in the two first years (July, November, January and April). Basal soil respiration was measured twice, in November 2004 and November 2005.

Mineral soil

Soil cores (0–12 cm) were collected from eight randomly selected points in each of the plots and at each sampling time, with a PVC core (50 mm diameter). The samples thus obtained were analyzed to determine chemical properties. The soil cores were transported to the laboratory, where composite samples from each plot were processed. Visible plant debris was removed and the soil samples were sieved (2 mm) and mixed to ensure homogeneity. Subsamples obtained for determining total C and N, available and total elements were dried at 40 ºC before analysis. Inorganic N and biological properties were analyzed in fresh soil samples.

In the soil, the pH was measured in H2O and 0.1 M KCl (soil: solution ratio 1:2.5) with a glass electrode. Total C, N and S were analyzed with a LECO Elemental analyzer. Extractable elements (Ca, Mg, K and P) were extracted by the Mehlich 3 procedure. This multielement extractant, which contains CH3COOH, NH4NO3, HNO3, NH4F and EDTA, is thought to remove the soil solution and readily exchangeable forms of metals. The elements in these two extracts were analyzed by ICP-OES.

Extractable soil mineral N and N mineralization rates were monitored throughout the two first years after treatment, between May 2004 (two months after treatments) and April 2006. Annual rates of N mineralization were measured by monthly in situ incubations (Raison et al. 1987). At each sampling time and in each of the plots, paired soil cores were collected, with a PVC core (50 mm diameter), from the upper 12 cm of the Ah1 horizon, at six random points. After removal of organic debris, one of the cores from each pair was sealed to prevent leaching, and replaced in its original site for incubation in the absence of plant uptake. After 30 days, the incubated soil cores were retrieved and analyzed to determine the final concentrations of ammonium and nitrate. Ammonium and NO3 were extracted with 2 M KCl and measured in a flux injection analyzer (FIAstar 5000). Monthly net N mineralization rates were calculated as the differences in mineral N content of the field-exposed and non-exposed soil core samples.

Basal respiration (the amount of CO2 evolved without addition of substrate) was determined after placing 70 g of fresh soil (three samples of each soil) in a glass container (1 L capacity) and adjusting the moisture content to 60% (by weight). The samples were incubated at 25 ºC for 10 days to allow the microbial processes to settle, and the emitted CO2 was measured.

Needle and litterfall

The effects of the treatments on forest vitality and nutrient cycling were also assessed by monitoring the litterfall and the amounts of nutrients returned in the litterfall, which were monitored for three years between 2005 and 2007. Needles from the upper third of the crown of at least 12 trees per plot were collected by aid of a hunting rifle. Litterfall was collected monthly in each of six litter traps (0.25 m2) located at random in the plot, and analyzed. Samples were pooled to provide two composite samples of foliage and litter per plot.

The oven-dried (60 ºC) samples of the plant material (needles and litter layer) were milled (0.25 mm) and digested with HNO3 in a microwave oven. Carbon, N and S in milled material were analyzed by combustion, with a Leco analyzer. The concentrations of P, K, Ca, Mg, Al, Fe, Mn, Zn and Cu in the digested plant extracts were measured by ICP-OES.

Nutrient resorption efficiency (NRE) was determined for the period prior to senescence, when the nutrient concentrations were relatively stable, as (Nm-Ns)/Nm, where Nm is the mean percentage of nutrient concentration in mature foliage prior senescence, and Ns is the concentration recorded in the last measurement made in autumn in senescent needles, collected in the litter traps (Oleksyn et al. 2003).

The litter accumulated in each plot was measured in November 2008. In each plot, six samples of litter layer (L + F), including leaves, twigs and reproductive organs, were collected at random with 0.3-m diameter sampling rings. All samples were weighed and subsamples were oven-dried at 65 ºC to convert fresh mass to dry mass.

The effects of the treatments on soil properties and vegetation were analyzed by analysis of variance. Data were assessed for homogeneity and normality. Temporal changes were explored by analysis of repeated measurements. Means were compared by Tukey’s test. Differences were considered significant at P < 0.05, for all parameters. Single and multiple-variable regression models were used to analyze the correlations between different parameters for each management method.

Tree growth

The trees were measured in April 2004, after thinning and before fertilizing, and again in April, 2008. Diameter at breast height (d) was measured at right angles. Total height (h) and height to the base crown (hbc) were measured with a hypsometer. Descriptive variables of each tree were recorded, e.g., mortality and crown breakage. A total of 795 trees were measured.

A portion of the trees displayed negative height increments due to crown breakage, and were not considered in the analysis. Because of this, top height was calculated from the plot specific height-diameter relationships, as the height assigned to the mean quadratic diameter of the 100 thickest trees per ha. Average top height in the site was 21.1 m (s.d. = 1.51 m).

Analysis of variance (ANOVA) was carried out by the PROC GLM procedure of SAS/STAT® (SAS Institute 2004), in order to test the effects of fertilization treatments on height and individual basal area growth, and covariates (initial diameter or height of each tree) were considered when necessary. The model used was:
$$ I{d_{{ij}}} = \mu + {F_i} + {B_j} + F*{B_{{ij}}} + {\varepsilon_{{ij}}} $$
(1)
where Idij is the 4-year diameter increment (or, alternatively, height increment), μ is the mean value for population, Fi is the effect of fertilization, Bj is the effect of the block, F*Bij is the interaction fertilization-block and εij is the error term.

For pairwise comparisons, the Tukey’s studentized range test was used. The LOGISTIC procedure of SAS STAT was used to test the effects of treatment on binomial variables. ANOVA was also used to evaluate the effect of fertilizer application on basal area growth at the stand level.

As the fertilizer was applied after thinning, the expected individual tree growth may have been enhanced by the reduction in competition due to the removal of trees. On average 56% of the trees were removed, and the basal area removed was 11.7 m2ha−1. The residual basal area was 26 m2ha−1 (sd = 4.3 m2 ha−1). The application of single tree growth models (Crecente 2008) enabled calculation of the predicted increment in tree basal area, considering the effect on growth of the reduced competition after thinning. This allowed a more direct exploration of the true effects of fertilization on growth, by use of the following model:
$$ ig = {\alpha_0} \cdot {d^{{{\alpha_1}}}} \cdot {G^{{{\alpha_2}}}} \cdot {e^{{{\alpha_3} \cdot t + {\alpha_4} \cdot BALMOD + {\alpha_5} \cdot BAR}}} $$
(2)
where ig is the annual increment in tree basal area g (cm2), d is breast height diameter (cm), G is stand basal area (m2/ha), t is age (years), BALMOD is the modified competition index developed by Shröder and Gadow (1999), which is equal to the relative basal area of larger trees divided by the relative spacing, BAR is the ratio tree basal area-stand basal area αi are parameters: α1 = 0,3674, α2 = 2,651, α3 = −0,754, α4 = −0,05207, α5 = −102.

Results

Nutrients in soils and foliage

The responses of the forest plots to the different treatments in terms of nutrient status are shown in Fig. 1. The changes in element concentrations in soil (Fig. 1a), green needles (Fig. 1b) and litterfall (Fig. 1c) are shown for the period 2004 to 2007. Nutrient resorption efficiency (NRE) is shown in the last column of the figure (Fig. 1d).
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-011-0794-y/MediaObjects/11104_2011_794_Fig1_HTML.gif
Fig. 1

Changes in soil pH and element concentrations in mineral soil (upper 12 cm), green needles and litterfall, and Nutrient Resorption Efficiency (NRE) throughout the study period. The green needle and litterfall data correspond to the samples collected in autumn. Changes over time were analyzed by analysis of repeated measurements. When this test revealed significant differences for the whole period, the monthly mean values were compared by Tukey’s test, at P < 0.05. Asterisks denote significant differences between the treatments compared within that month

None of the treatments led to clear changes in soil mineral pH or exchangeable Al saturation (the latter data are not shown). Although the treatments did not lead to changes in N in green needles, the concentrations of litter N decreased significantly relative to those in the untreated plots, possibly because of the decreased availability of soil mineral N (see below). Nitrogen was the only element for which the resorption values were increased by the treatments.

The three treatments led to increases in Ca in soil, in green needles and in litterfall in the three first years. Despite the higher input of Ca in treatment CaP (493 kg ha−1) than in the MWA or MWAP treatments (186 kg ha−1), the concentrations of this element were not higher in green needles or litterfall. As expected, the concentrations of Ca were higher in the litterfall than in green needles, leading to negative resorption values.

Moderate increases in Mg concentrations in soil and litterfall were observed in response to the three treatments. This was also observed in treatment CaP, even though no Mg was added. This effect may be due to the replacement of this element by Ca at the soil exchange sites. In accordance with the higher supplies, the Mg resorption values were significantly lower in these plots than in the untreated plots.

The application of wood ash (treatments MWA and MWAP) led to higher K concentrations in soil and litterfall, but not in green needles (a slight decreasing trend was observed). Resorption of this element decreased during the first years.

Soil available P increased in the plots to which mineral P fertilizer was added (treatments CaP and MWAP). The concentrations of P in green needles and litterfall were also higher in these plots during the first years. In relation to the higher supply of this element, lower resorption values were recorded in these plots. Although the amounts of P added in the CaP and MWAP treatments were very similar, the concentrations of soil available P and litterfall P were lower in the plots treated with MWAP, which suggests that the wood ash may have reduced the availability of this element in the soil.

Mineralization of N and C

The average concentrations of NH4+ and NO3 in the untreated soil were 3.5 mg kg−1 and 3.1 mg kg−1, respectively (Fig. 2). All treatments led to large decreases in extractable NO3. In the plots to which MWA was added, the concentrations of extractable NH4+ were slightly lower than in the untreated plots.
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-011-0794-y/MediaObjects/11104_2011_794_Fig2_HTML.gif
Fig. 2

Concentrations of NO3-N and NH4+-N in soil (0–15 cm depth) in the different plots. Values given are means and standard errors of four measurements. Asterisks denote significant differences between the treatments compared within that month

In the untreated soil, the annual mineralization rate for the two years studied was 28 kg ha−1yr−1(Fig. 3), and most of this was accounted for by nitrification (19 kg ha−1yr−1). All treatments led to significant decreases in soil N mineralization. Treatment CaP had a stronger effect on ammonification, whereas the addition of MWA led to important decreases in nitrification, with the effect being more notable in the second year. For the two years of the soil incubation, the amount of N mineralized was 20–35% lower than in the untreated soils.
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-011-0794-y/MediaObjects/11104_2011_794_Fig3_HTML.gif
Fig. 3

Net monthly N mineralization in soil after harvesting and site preparation (0–15 cm depth). Values given are means of four measurements. Different letters indicate significant differences between means according to the Tukey test

Furthermore, soil respiration (Fig. 3) was studied in the upper 10 cm of the mineral soil in 2004 and 2005, four and 20 months after the treatments, respectively. The analysis revealed slight increases in soil respiration rates in the soils treated with MWAP relative to those in the untreated plots, during the first months after treatment.

Litterfall and nutrient return

Litterfall and the amounts of nutrients returned in the litterfall throughout the three years after treatment is shown in Figs. 4 and 5. The annual litterfall in the untreated plots was 3.7 Mg ha−1 yr−1, with maximum values in spring and autumn. Litter production increased significantly in the plots treated with MWAP and CaP in the seasons with highest litterfall, and differences were still observed in spring three years after treatments. Increases of up to 17% were recorded in the plots treated with MWAP and CaP during the three year period of study.
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-011-0794-y/MediaObjects/11104_2011_794_Fig4_HTML.gif
Fig. 4

Cumulative amounts of litterfall and nutrients returned by litterfall throughout the period December 2004–November 2007. Asterisks denote significant differences between the treatments compared within that month

https://static-content.springer.com/image/art%3A10.1007%2Fs11104-011-0794-y/MediaObjects/11104_2011_794_Fig5_HTML.gif
Fig. 5

Total amounts of litterfall and nutrient returned by litterfall throughout the period December 2004–November 2007. Different letters indicate significant differences between means according to the Tukey test

The amounts of elements returned by litterfall throughout the three years are shown in Fig. 4, whereas the total amounts returned for the whole period are shown in Fig. 5. The different litterfall productions in the plots were due to differences in litterfall production and/or to differences in litter nutrient concentrations. With respect to the C and MWA plots, treatments CaP and MWAP showed higher returns of P, Ca and Mg. The amounts of P returned were similar in these treatments (17% of the P added), around 3 kg ha-1 higher than in the untreated plots. In the case of Ca, the amounts returned were 8 kg ha-1 higher than in the untreated plots, and represented 11 and 4% of the amounts added in treatments MWAP and CaP, respectively. However, the application of MWAP led to higher returns of K and Mg (13 and 34% of the amounts added, respectively), reflecting the higher inputs of these elements.

In relation to the higher litter production with respect to the untreated plots (r = 0.95, p < 0.01), more litter accumulated in stands treated with CaP (39 Mg ha−1), MWA (37 Mg ha-1) and, particularly in those treated with MWAP (49 Mg ha−1), with respect to the untreated plot (34 Mg ha−1).

Tree growth

Fertilization after thinning significantly affected (p = 0.0035) stand basal area growth. MWAP increased the 4-year stand basal area increment in the control (5.2 m2 ha−1) by 40%, slightly more than the 33% the increment produced by CaP and much higher than the 13% increase produced by MWA.

The initial diameter was a highly significant (p < 0.0001) factor in explaining the 4-year diameter growth. The effects of fertilization and block were also significant, as was the interaction (Table 3). A detailed analysis indicated that this involved a decrease in fertilization effects in the blocks with higher growth, and some differences in the ranking of the treatments CaP and MWAP in the blocks with poorest growth. Significant differences were not observed in diameter growth between MWAP, MWA and the control plots.
Table 3

Results of covariance analysis for 4-years individual diameter and height growth, considering all the trees and a subsample of the 250 thickest trees per ha. Different letters indicate significant differences between means according to the Tukey test

Individual diameter growth (cm)

Individual height growth (m)

Covariance analysis

Ranking of treatments

LS means

Covariance analysis

Ranking of treatments

LS means

Analysis for all the trees

n = 795

CaP

4.19 a

n = 780

MWAP

3.44 a

F: p = 0.0028

MWAP

3.82 ab

F: p = 0.0359

MWA

3.19 ab

B: p = 0.0002

MWA

3.66 b

B: p = 0.007

CaP

3.16 ab

F*B: 0.0011

C

3.53 b

F*B: p < 0.0001

C

3.06 b

250 thickest trees per ha

n = 480

CaP

4.89 a

n = 478

MWAP

3.62 a

F: p = 0.0368

MWP

4.52 ab

F: p = 0.0019

MWA

3.41 a

B: p = 0.0425

MWA

4.44 b

B: p = 0.0015

CaP

3.12 b

F*B: 0.3319

C

4.19 b

F*B: p = 0.0038

C

3.00 b

ns not significant. B Block. F Fertilization

The influence of initial height in explaining height growth was not significant (p = 0.0574). The effect of fertilization was significant, although so was the interaction with block (Table 3), and further analysis did not clarify which treatments enhanced height growth.

Further analysis was carried out by considering the 250 thickest trees per ha, i.e. those most likely to form the future crop trees. The results were quite similar to those already obtained (Table 3).

Treatments did not have any significant effects on tree mortality, stem breakage (due to wind), or crown length.

The relationship between observed and predicted increment in tree basal area for each treatment is shown in Fig. 6. As the predicted values are those expected after the reduction in tree competition due to thinning, these results provide a clear insight into the synergy between thinning and fertilization. The treatments that were most useful for enhancing diameter growth after thinning were those in which mineral P was applied. Individual growth in the control plots closely followed the expected trend (with a difference of less than 5%). The observed values for the MWA plots were slightly higher than the predicted values. Correction factors of 1.135 and 1.153 were required to estimate growth for treatments in which P was added.
https://static-content.springer.com/image/art%3A10.1007%2Fs11104-011-0794-y/MediaObjects/11104_2011_794_Fig6_HTML.gif
Fig. 6

Observed against predicted 4-years individual-tree basal area increments for each treatment, obtained after applying a competence-specific model. Solid lines represent a non-intercept linear equation fitted to each scatter plot of data and the dashed line is the diagonal

Discussion

pH, Ca, Mg and K in soil and plant samples

Application of MWA at the dose used did not affect the soil pH, and the increases in available nutrients were moderate because of the relatively low content of easily dissolved nutrients in the ash. The responses were less intense than those recorded in previous studies with fly wood ash, from which release of elements is faster (Augusto et al. 2008; Solla-Gullón et al. 2006; Pérez-Cruzado et al. 2010).

Only slight changes in the concentrations of most nutrients in green needles were observed, as also been observed in other studies (e.g. Helmisaari et al. 2009). The lack of this response is attributed to the dilution of nutrients due to the enhanced needle production resulting from the higher nutrient availability (Solla-Gullón et al. 2008). In the present study, the higher litterfall production confirmed that the treatments increased needle production. The higher concentrations of P, Ca, Mg and K in litter and the lower efficiency of resorption of these elements after treatments also reflect higher uptake by plants. These results suggest that litterfall monitoring may be more useful than analysis of green needles for evaluating effects on nutrient status.

Phosphorus

The MWA used in this study contained low amounts of P in comparison with fly wood ash (Solla-Gullón et al. 2006) or ash from other crops (Schiemenz and Eichler-Lobermann 2010). For this reason, no evidence of a higher supply of P (soil, needle, litterfall) after a single application of ash (MWA treatment) was expected. Nevertheless, some studies have shown improvements in P supply in soils treated with charcoal. It is thought that charcoal favours mycorrhizal associations as well as the development of heterotrophic phosphate solubilizing microorganisms (Warnock et al. 2007; Makoto et al. 2010). It is possible that the amount of ash used in the present experiment was not sufficient to promote ectomycorrhizal formation.

The increases in soil available P in response to treatments including supplementary mineral P (treatments CaP and MWAP) were moderate, possibly due to sorption of P by clay and Fe-Al oxides as a consequence of the low soil pH. Nevertheless, the significant increases in P concentrations in plants, along with the greater amounts of P returned by litterfall, suggest that addition of mineral P may compensate the lack of P in MWA.

Interestingly, although almost the same amount of P fertilizer was added in both treatments (CaP and MWAP), the responses were lower for MWAP (Fig. 1). This may be due to the presence of charcoal in the soil after treatment. Charcoal has been shown to retain P and reduce the dissolved P leaching (Laird et al. 2010). It is possible that charcoal can absorb the inorganic P added as poorly soluble phosphates of Ca and Mg.

Nitrogen

The rate of N mineralization recorded in this soil (28 kg N ha−1 yr−1) was close to the input of N via litterfall (29 kg N ha−1 yr−1). The data are consistent with the results reported by Pérez-Batallón et al. (2001) and Fernández et al. (2009) for mature Pinus radiata plantations in the region.

The results of the field incubations indicated that all the treatments—MWA, MWAP and CaP—decreased the N mineralization from the soil. The data are consistent with that obtained in a previous study carried out after addition of fly wood ash in the same region (Solla-Gullón et al. 2006). In the stands treated with MWA, the decreased nitrification rates, as well as the low extractable NO3−1, indicate there is no risk of NO3−1 leaching as a consequence of any treatment, unlike what has been observed in other environments (Ball et al. 2010). On the other hand, the lower availability of N in the soils treated with wood ash led to decreased N supply, as revealed by the lower concentrations of N in litterfall and the higher resorption efficiencies. In these plantations, which are not limited by N, this effect probably did not affect production. Under other conditions, decreased N availability could induce N deficiency in plants. Addition of wood ash and of Ca(OH)2 may have enhanced temporal N immobilization via increased microbial activity as a result of the higher soil nutrient availability and/or the greater release of DOC (Corre et al. 2003; Jokinen et al. 2006), and this was confirmed by the higher soil respiration rate. In addition, in the treatments with wood ash, the high C:N ratio in the charcoal, may also have promoted immobilization of N. Decreases in N mineralization due to the presence of charcoal have also been found elsewhere (Steiner et al. 2008). In addition to these possible mechanisms, mineral N may also be immobilized by chemical reactions between mineral and organic matter (Dail et al. 2001). However, under certain conditions the charcoal remaining after wildfires of low-moderate intensity has been shown to enhance nitrification (Ball et al. 2010). However, the latter effect has been attributed to the charcoal-induced suppression of phenols that inhibit nitrifying bacteria.

Litterfall, nutrient return and nutrient resorption

The annual amounts of litterfall and nutrient returned by litterfall recorded in this study were consistent with those recorded in other older Pinus radiata plantations in the region (Ouro et al. 2001).

The increased litterfall production and litterfall nutrient concentrations in response to treatments MWAP and CaP indicate an early response of trees in terms of biomass production. This shows that the ecosystem is sensitive to the higher nutrient supplies after these treatments. Thus, the higher returns of K and Mg found in treatment MWAP confirmed that wood ash supplies greater amounts of these elements to forests. On the contrary, the lack of response to the MWA without addition of mineral fertilizer confirms that, at the dose used, the supply of P was not sufficient to correct this limitation.

Different studies have also shown enhanced litter production following mineral fertilization (Campo et al. 2007; Kaspari et al. 2008, Albaugh et al. 2008). The higher production in response to MWA also led to higher litter accumulation. Since the accumulation of litter was proportional to the litterfall production, this suggests that the litter decomposition rates were not affected by the treatments. Application of MWA can alter soil biological activity, through different mechanisms, as a consequence of increased soil pH (Corre et al. 2003; Jokinen et al. 2006) and the presence of charcoal (Liang et al. 2010; Wardle et al. 2008). The results of the present study show that at the dose used, which was much lower than in the latter studies, the MWA did not cause increased decomposition of litter.

Increased nutrient concentrations in litterfall after fertilization of soils have been described in other types of ecosystems (Sardans et al. 2005; Campo et al. 2007), and were probably due to the lower resorption efficiency, which is considered an adaptation of the trees to the increased nutrient availability (Oleksyn et al. 2003). However, the relation between resorption efficiency and nutrient availability is not always evident, because of the involvement of other environmental and genetic factors (Oleksyn et al. 2003; Blanco et al. 2009).

Tree growth

The results of tree growth were consistent with the improved nutrient status brought about by treatments CaP and MWAP. The approach used to analyze the tree response in terms of growth was particularly sound for thinned stands, as the reduced competition between trees was able to be distinguished from the effects of fertilization. Snowdon and Waring (1990) found it useful to use a model including a competition index to describe differences in basal area growth between types of thinning. In the present study, as different thinning intensities were not applied, a model was used as a baseline to compare tree growth. Nevertheless, the multiplicative effect on growth must be studied over a longer period to determine whether the fertilization response is a type 1 response (short term only) or type 2 response (long term, definitively increasing the site productivity) (Snowdon 2002).

Evaluation of the effect on growth was carried out four years after treatment, which is probably close to the age of maximum response of fertilization in individual diameter growth in mid-rotation plantations (Hynynen et al. 1998).

Ash can be applied as a fertilizer at different stages of stand development, but is usually applied before the age at which maximum volume growth occurs (10 years before the age of the plantation under study here). However, increases in growth after wood ash application have not always been observed (Jacobson 2003; Saarsalmi et al. 2004).

The present study shows that fertilization at later stages of stand development, i.e. after the last thinning, may still have significant effects on growth. Such effects may be lower than those produced by fertilizer applied after the first thinning, but the wood is more valuable and the period before final harvest is shorter, leading to maximal returns for the investment in fertilizer (Turner et al. 1992). In the present study, the economic evaluation revealed positive results for treatments combined with superphosphates, but not for ash alone. Calculations were made assuming no additional absolute responses after 4 years post thinning and a rotation age of 33 years, with a discount rate of 0.03.

Conclusions

The aim of this study was to investigate the mechanisms involved in the response of forest ecosystems to the application of MWA, a complex mixture of fly ash, charcoal and unburned material. Although this type of wood ash contains valuable amounts of readily available elements (K, Ca and Mg), there is a lack of P and N in the product. In addition, charcoal may have important effects on the dynamics of these elements.

The results showed that the lack of P in MWA can be compensated by supplementation with mineral P fertilizer. However, the abundant charcoal in MWA may have temporarily reduced the solubility of mineral P. Because of the higher microbial activity in response to the better nutrient availability, along with the effect of the very high C/N ratio in the wood ash, the application of MWA leads to decreased N mineralization and soil mineral N, which should be taken into account in N-limited ecosystems.

Monitoring of litterfall production and nutrient concentration was successful in evaluating the early response of trees in terms of nutritional status. Thus, the changes in nutrient supply were reflected in changes in nutrient resorption efficiencies. The use of density-dependent single tree increment models confirmed an increase in the value of the trees as a consequence of the better nutrient status brought about by supplementation of the MWA with mineral P fertilizer.

Acknowledgements

Funding for this research was provided by the Spanish Ministry of Education and Science, and Financiera Maderera S. A. (FINSA). We are grateful to Sergio Blanco, Ramiro García and Victor Torrado (FINSA) for their help in carrying out the project. Analyses of the soils and plant material were carried out by Verónica Piñeiro and Montse Gómez (RIAIDT-University of Santiago de Compostela).

Copyright information

© Springer Science+Business Media B.V. 2011