Advertisement

Agroforestry Systems

, Volume 92, Issue 2, pp 301–310 | Cite as

Carbon fractions as indicators of organic matter dynamics in chestnut orchards under different soil management practices

  • Olga Borges
  • Fernando Raimundo
  • João Coutinho
  • Berta Gonçalves
  • Ivo Oliveira
  • Afonso Martins
  • Manuel Madeira
Article

Abstract

Several studies have emphasized the negative impact of the conventional soil management (CT) system on productivity and sustainability of chestnut orchards (Castanea sativa Mill.) when compared to no-tillage with grass cover (NT). However, scarce information is available regarding the effects of these soil management systems on soil organic matter (SOM) dynamics and soil quality. SOM or soil organic carbon is a key component of soil quality and has different fractions with different lability, namely, organic C (POC), active C (AC) and hot-water extractable carbon (HWC). These are considered as indicators of changes in management-induced soil quality. Thus, a study was carried out to evaluate the effects of NT and CT systems applied in the chestnut orchards on: (i) total amount of soil organic C (TOC), including C from both organic and mineral layers; (ii) soil organic C concentration of mineral horizons (OC); (iii) labile soil organic fractions (POC, AC, HWC); (iv) and soil mineral-associated C. The study was developed in two 30-year old chestnut orchards located in Northeast Portugal, that were kept under different soil management systems (NT or CT) during the preceding 17 years. Soil samples were taken at 0–10 and 10–20 cm soil depth. No significant differences in OC concentration were observed between NT and CT, while TOC was significantly higher in NT than in CT (22.54 and 12.17 Mg/ha or 34.16 and 22.90 Mg/ha, considering the organic layer plus mineral layers at 0–10 and 0–20 cm depth (set of two depths). The NT practice led to significantly higher concentration of labile C fractions (POC, AC and HWC) than CT at 0–10 cm soil depth. These results indicate that measurement of labile soil organic C fractions, such as POC, AC and HWC, may provide a sensitive and consistent indication of changes in soil C and SOM dynamics in response to soil management practices. Overall, NT seems to ensure better soil quality than CT in chestnut orchards under Mediterranean climate conditions.

Keywords

Castanea sativa Labile soil organic carbon Soil quality Soil tillage 

Introduction

In the last decades, soil quality has received particular attention, focused on the impact of management practices on soil quality and identification of soil parameters or attributes that can be used as indicators of such quality (Duval et al. 2013). Soil organic matter (SOM) is considered a key component of any terrestrial ecosystem and it is probably the most largely acknowledged indicator of soil quality, as it influences soil biological, physical and chemical properties and processes, and is strongly associated to sustainable agroecosystems (Chen et al. 2009; Haynes 2005). SOM includes a continuum of materials ranging from highly decomposable material, the labile organic C, to very recalcitrant, which corresponds to the stabilized organic C (Haynes 2005), having no definite chemical composition and it is usually measured as soil organic C (SOC) (Duval et al. 2013; Weil et al. 2003).

Labile soil organic C fractions, like particulate organic C (POC), hot-water extractable C (HWC) and active C (AC), that respond in a short time period (<10 years) to the management-induced alterations, have been used as early indicators of changes in SOM, since changes in total SOM content are difficult to detect, due to the relatively large quantity of background organic matter already present, and only are usually demonstrated in the long-term experiments (e.g., >25 years) (Chen et al. 2009; Haynes 2005; Melero et al. 2009; Weil et al. 2003). It is known that the soil’s ability to act as a C sink is linked to several factors, including geographic location, environmental factors, and management practices (Jose 2009), and thus, the potential for the accumulation of C in a given agroforestry is affected by soil structure and aggregation (Nair et al. 2010). These carbon levels are, in temperate regions, improved by agroforestry systems, when compared to other land-use systems (Paudel et al. 2011). In fact, agroforestry systems are identified with an increase of the resource-use efficiency of the system (Eichhorn et al. 2006), providing market (food, wood products, and fodder) and nonmarket goods and services (soil conservation, water and air quality improvement, biodiversity and scenic beauty) (Alavalapati et al. 2004). Furthermore, no-tillage systems with grass cover decrease the loss of SOM by erosion, improves water infiltration and reduces runoff, increases soil nutrient retention and builds SOM; in contrast, tillage increases outputs of SOM trough physical disturbance of soil structure (Nyamadzawo et al. 2009) increasing the loss of soil organic carbon (Lalitha and Kumar 2016). Harmful effect of conventional tillage has been found in chestnut orchards, with lower productions, higher risk of erosion and enhanced probability of ink disease (Phytophthora cinnamomi) (Martins et al. 2010, 2011). On the contrary, no-tillage with grass cover controlled by grazing with sheep and cutting in the late spring (NT) increased fruit and mushrooms production, including edible mushrooms, improved fruit quality parameters, did not affect soil water content and leaf water potential and lead to higher productivity of the system (Martins et al. 2010, 2011; Raimundo 2003). This is of extreme importance, as the chestnut for fruit production is an important crop in the Mediterranean area, for which Portugal represents 42% of the European chestnut area (Martins et al. 2011), and is one of the most significant sources of income for rural areas. As a result, farmers have intensified management practices in order to increase productivity of chestnut orchards. However, scarce information is available about the effects of soil management practices on soil C sequestration and on the dynamics of labile soil organic C fractions in chestnut orchards under Mediterranean conditions. Therefore, a study was carried out to assess the effects of medium-term (17 years) soil management systems, NT and CT, on: (i) soil organic C of mineral soil layers (OC); (ii) total amount of soil organic C (TOC), which includes organic C of organic layer plus organic C of mineral layers up to 20 cm depth; (iii) labile organic C fractions, POC, AC and HWC; (iv) mineral-associated C (MAC); and (v) relationships between OC, labile C fractions and MAC. It was hypothesized that NT increases the concentration of SOC and the proportion of labile organic C fractions and that these fractions are reliable indicators of soil quality.

Materials and methods

Study site and soil management practices

The studied orchards are located at Terroso (municipality of Bragança), Montesinho Natural Park, northeast Portugal (latitude 41°52′52″–41°52′54″N; longitude 6°50′11″–6°50′13″W; altitude 906 m a.s.l.). The climate is of Mediterranean type with warm dry summers and cool and wet winters. Considering the available climate data for the meteorological station at Bragança (10 km far from the studied orchards) for the period 1971–2000, the mean annual rainfall was 758.3 mm, mainly concentrated from October to May (84%), and the mean annual temperature was 12.3 °C. The monthly mean air temperature ranged from 4.4 °C in winter (January) to 21.3 °C in summer (July) (IPMA 2013). The area corresponds to levelled land (Bragança tableland) and predominant slope gradient is <2%. Soils are mostly Dystric Cambisols (Agroconsultores and COBA 1991) developed over schist formations, with sandy loam to loam texture. As shown in Table 1, the soils are mostly acidic (pH values in H2O ranging from 5.10 to 5.35), with SOM concentration close to 2.5 and 1.8% respectively at the 0–10 and 10–20 cm layers. Available P2O5 and K2O concentrations were respectively medium and very high at the surface top soil layer. The effective cation exchange capacity (ECEC) values were low, ranging from 3.33 to 4.00 cmolc kg−1 in the 0–10 cm soil layer, respectively under CT and NT systems, and from 3.62 to 3.73 cmolc kg−1 in the 10–20 cm depth, respectively in NT and CT. The base saturation percentage at soil pH, in each layer was higher in the NT than in CV systems.
Table 1

Chemical characteristics in the 0–10 and 10–20 cm soil layers from the no-tillage (NT) and conventional tillage (CT) systems

Depth (cm)

Tillage

SOM (%)

P2O5

K2O

pH (H2O)

Ca2+

Mg2+

K+

Na+

EA

ECEC

mg kg−1

cmolc kg−1

0–10

NT

2.64

82.0

498.0

5.35

1.92

0.81

0.16

0.15

0.96

4.00

CT

2.48

74.5

345.0

5.25

1.22

0.56

0.14

0.11

1.30

3.33

10–20

NT

1.78

49.0

217.0

5.30

1.35

0.60

0.12

0.14

1.41

3.62

CT

1.88

43.0

176.0

5.10

0.92

0.54

0.20

0.09

1.98

3.73

Values are means (n = 16); no statistically significant differences were found between different soil management practices (α > 0.05), for each depth

SOM soil organic matter, EA exchangeable acidity, ECEC effective cation exchange capacity

Two adjacent 30-year-old Castanea sativa Mill. orchards (Longal cultivar), in similar geology, topography, soil type and texture (loam), with a tree spacing of 12 × 12 m, which only differed in the soil management system for the last 17 years (1995–2012), were selected. Two soil management systems were assessed: conventional soil tillage (CT) using a tine cultivator, down to 15–20 cm depth, twice a year, that occurred in winter to incorporate organic residues and fertilizers and in late spring to control weeds; no-tillage (NT)—maintenance of grass cover (spontaneous herbaceous vegetation) controlled by route grazing with sheep and cutting with heavy-duty mulcher (rotor with hammers) in spring and before fruit harvest. This management treatment was installed 17 years ago in a chestnut orchard previously managed as CT. Grass cover was mainly composed of Vulpia bromoides (L.) Gray and Agrostis castellana Boiss. et Reut. covering 75–90% of the soil surface.

Soil sampling

Sampling was conducted throughout the year 2012. Inside each of chestnut orchard area, presenting each of the management systems (CT or NT) an area of 300 m × 200 m was delimited, and within it 15 plots (70 m × 60 m) were established. To alleviate possible pseudoreplication problems (Stamps and Linit 1999), four of them were randomly selected. Eight trees were randomly chosen within each sampling plot.

In the NT system, organic litter layer samples were collected beneath tree crown and in the open (two randomly selected trees per sampling plot) with a 25 cm diameter stainless ring, obtaining a total of 16 samples. These samples were dried at 65 °C until constant weight and ground through a 1-mm screen.

In both treatments, soil mineral samples were collected at two depths, 0–10 and 10–20 cm, under and out of tree canopy. In total, 128 samples for each system were taken (8 trees × 2 positions × 2 depths × 4 plots). In each selected tree of each sampling plot, one sample was taken both beneath tree crown (BC) and in the open (OP), and samples of each position (BC or OP) were combined to give two composite samples; that is, 16 composite samples were obtained for each soil depth of each treatment system. In the NT, soil samples were taken after removing the surface organic layer. The mineral samples were dried at 45 °C, then passed through a 2 mm sieve and stored at room temperature for further analyses.

In half of the sampled locations and for each layer, undisturbed soil samples were also taken for bulk density measurement using the core method (Blake and Hartge 1986).

Soil analyses

Soil organic matter concentration from organic layer was determined by ignition loss after heating at 450 °C for 6 h. An assumption of 58% of organic C content in organic matter was followed. In the mineral soil layers, organic C concentration (OC) was determined by dry combustion using a Skalar PrimacsSCN analyzer. Total organic C (TOC) in each treatment up to 20 cm depth was calculated as the sum of the amount of organic C of mineral and organic layers and expressed in Mg/ha. Total N was determined by the Kjeldahl method (Houba et al. 1986) and available P and K by the Egnér–Riehm test (Egnér et al. 1960), which were quantified by molecular absorption spectrophotometry and by flame emission spectrophotometry, respectively.

Exchangeable acidity was extracted with KCl 1 M solution and quantified by titration with NaOH 0.05 M (Thomas 1982). In the mineral layers, exchangeable bases were extracted by 1 M ammonium acetate, adjusted to pH 7 (Houba et al. 1986), and quantified by atomic absorption spectrophotometry for Ca and Mg and by flame emission spectrophotometry for K and Na. The soil pH was measured in a 1:2.5 soil/water suspension.

Labile soil organic carbon fractions

Active C (AC) was determined as described by Weil et al. (2003). Briefly, 5 g of soil reacted with 20 mL of 0.02 M KMnO4 in 1 M CaCl2 (pH 7.2). The soil suspension was shaken on a reciprocating shaker for 2 min and then allowed to settle for 10 min. A 200 µL aliquot of the solution supernatant was diluted to 10 mL of deionised water and absorbance was measure on a spectrophotometer (Hélios Omega UV–VIS) at 550 nm.

Particulate organic C (POC) was determined using the procedure of Cambardella and Elliott (1992). Twenty-five grams of soil sample were shaken with sodium hexametaphosphate to disperse soil aggregates. The soil suspension was poured through a 53 µm sieve and rinsed with deionised water until the clay and silt-size fractions were completely removed. The organic C was determined by the wet oxidation method (De Leenheer and Van Hove 1958).

Hot-water extractable C (HWC) was performed according to Ghani et al. (2003). Ten grams of soil were dispersed in 40 mL of deionised water, using 50 ml polypropylene centrifuge tubes and shaking on a horizontal shaker at 30 rpm for 30 min. Afterwards, the tubes were left for 16 h in an 80 °C hot-water bath, followed by centrifugation for 20 min at 8000 rpm. The supernatant was vacuum filtered through 0.45 μm cellulose acetate membrane filters. Extracts were analysed for total organic C on a FormacsHT C analyzer.

Soil mineral-associated carbon

The mineral-associated C (MAC) was calculated by the difference between the concentration of OC and the concentration of POC (Bayer et al. 2004).

Statistical analyses

Statistical data analyses were carried out using JMP 5.0.1.2. (The Statistical Discovery Software, Copyright© 1989–2003 SAS Institute Inc.) for Windows. All data in tables are presented as means. A one-way ANOVA was performed in order to compare the effects of soil management practices and sampling depths. Significant differences between soil management practices and between depths were determined by a Tukey–Kramer HSD test at α < 0.05. A correlation matrix of OC and C fractions was based on Pearson correlation coefficients (α < 0.001, α < 0.01 and α < 0.05). A principal component analysis (PCA) was applied for reducing the number of variables (all analyzed parameters) to a reduced number of new derived variables (principal component or factors) that summarize the original information, that is, C fractions. A linear discriminant analysis (LDA) was used as a supervised technique to classify soils according to their C fraction concentration. A stepwise technique, with Wilk’s lambda method and the usual probabilities of F (3.84 to enter and 2.71 to remove), was used for the selection of variables (Maroco 2003). Using this approach, it is possible to identify significant variables among all those in study. To verify which canonical discriminant functions were significant, the Wilks’ lambda test was applied.

Results and discussion

Soil organic C under or outside the chestnut’ canopy

Although soil sampling was performed both beneath and outside tree crown, no significant differences for OC, TOC and C fractions were found between those two sampling positions (data not shown). This pattern may be associated with the low open area within the total orchard, due to the great volume of chestnut tree canopy, or with the effect of the distribution of the residues, resulting for tillage or vegetation crushing. Thus, the sampling location factor (under or outside of the canopy area) was not considered in the following discussion.

Soil organic C in mineral soil layers

Tillage is known to induce a decrease in SOC content through mineralization and removal of crop and animal biomass, although these effects can be reversed by the adoption of restorative land practices such as agroforestry (Lal 2003). However, in the present study, soil OC concentration was not a sensitive and reliable indicator for assessing the impact of CT and NT systems on soil quality in chestnut orchards. The results showed that the latter system failed to enhance significantly differences in OC concentration when compared to the CT system, at both soil depths (Table 2). In fact, similar OC concentrations were observed in the NT and CT systems. Changes in OC concentration in response to soil management practices may be difficult to detect because of the generally high background and natural soil variability (Blair et al. 1995). Also, Haynes (2005) reported that changes on soil quality caused by different soil management systems occur much more slowly in OC than in labile fractions. Therefore, strong changes in soil are better revealed by organic C fractions, that represent a high percentage of the total soil organic C, whereas slow dynamics changes are better discriminated by labile soil organic C fractions (Oyonarte et al. 2007).
Table 2

Concentrations of organic C (OC), active C (AC), particulate organic C (POC), hot-water extractable C (HWC) and mineral-associated C (MAC) in mineral soil layers, and total amount of organic C (TOC) up to 20 cm depth, in conventional tillage (CT) and no-tillage (NT) systems

Tillage

Depth (cm)

OC (g kg−1)

TOC (Mg ha−1)

AC

POC

HWC

MAC

g kg−1

NT

0–10

15.31 ± 3.82A

22.54 ± 1.44Aa

0.34 ± 0.09Aa

4.57 ± 1.48Aa

0.66 ± 0.21Aa

10.24 ± 2.37A

10–20

9.77 ± 1.34B

11.62 ± 0.72B

0.16 ± 0.04B

1.58 ± 0.31B

0.33 ± 0.05B

8.27 ± 1.54B

0–20

12.54 ± 3.98

34.16 ± 1.62a

0.25 ± 0.11a

3.07 ± 1.84

0.49 ± 0.23

9.26 ± 2.21

CT

0–10

14.80 ± 1.85A

12.17 ± 0.52Ab

0.25 ± 0.04Ab

3.22 ± 0.90Ab

0.50 ± 0.05Bb

11.18 ± 2.21A

10–20

10.89 ± 1.83B

10.42 ± 0.45B

0.15 ± 0.06B

1.44 ± 0.24B

0.31 ± 0.06A

9.34 ± 1.58B

0–20

12.85 ± 2.69

22.60 ± 0.72b

0.20 ± 0.07b

2.33 ± 1.11

0.41 ± 0.11

10.26 ± 2.11

Values are mean ± SD (n = 16); within a column, different small letters indicate significant differences between the same depth for different soil management practices; in a column, different capital letters indicate significant differences between depths, within each soil management practices (α < 0.05)

Both soil management systems showed significantly higher OC concentration in the surface (0–10 cm) than in the subsurface (10–20 cm) soil layer. This difference was higher under NT (33%), than in the CT (24%) (Table 2). This variation of OC content between soil layers may be associated with the fact that tillage distributes C inputs more evenly throughout the ploughing layer, as reported by Six et al. (2002).

Changes in C stock

Results show a considerable effect of soil management systems on the total amount of organic C (TOC), a trend that has been previously reported by other authors (Raimundo 2003). The CT system resulted in a significant decrease of TOC, but only in the organic layer plus the mineral layer at 0–10 cm depth (Table 2). In fact, in this layer, the TOC in the NT system was almost the double of that measured in the CT. When analyzing the two combined layers (0–20 cm), significant differences were found between TOC amounts in the two soil management systems (Table 2). The decrease on the organic carbon recorded in the CT systems is probably due to exposure of the organic carbon to aeration environment, changing the soil microbial community structure and function resulting in high decomposition rate and has been recorded elsewhere (Lalitha and Kumar 2016). Furthermore, these significant differences indicate, that orchards located in the same edapho-climatic condition, but under different systems, are able to sequester significant different amounts of C per area (Weerasekara et al. 2016). These results may highlight the importance of the organic residues of chestnut and of grass cover in providing organic matter to the system, as well as the negative effect of soil perturbation by tillage. However, the complex dynamics of TOC in chestnut orchards cannot be overlooked. In fact, in the same region Raimundo (2003) showed significant differences in TOC amount after 4 years of NT, compared to CT, but, after 7 years, no significant differences were recorded. Grass covers contribute to the maintenance of soil structure, since they have a positive effect on soil physical protection by intercepting and reducing the impact of rain drops (Haynes 1980). Also, other authors emphasized the importance of soil cover on soil quality, regarding SOM (Hernández et al. 2005), biological properties (Moreno et al. 2009) and plant growth (Haynes 1980). When analyzing the variation of TOC with soil depth (0–10 cm layer and 10–20 cm layer) the trend was similar to that observed for the OC concentration (Table 2), with significantly higher values of TOC at the superficial soil layer compared to subsurface layer, in both systems.

Labile soil organic C fractions

Considering the obtained results for the labile soil organic C fractions, at the first soil layer (0–10 cm), significant differences between soil management systems were observed (Table 2), with higher concentrations of the three analyzed C fractions in the NT soils. The concentrations of each C fraction present in the second soil layer (10–20 cm) were found to be similar between NT and CT systems. However, significant differences were observed when comparing the concentration of AC in the two combined depths (0–20 cm). The influence of soil depths has been recorded in other agroforestry systems, and is an indicator of the importance of this system by reducing greenhouse gases emission, through the storage of organic carbon, protected in the soils and in high amount (Ramesh et al. 2015).

The greatest difference regarding labile soil organic C was observed for the POC fraction, with about 29.5% less organic C in the CT than in NT, at 0–10 cm depth. The depletion of this fraction may compromise the other C fractions, as POC is considered the precursor for formation of other labile soil organic C fractions and it is the main pathway through which C is returned to the soil from above-ground residues and root turnover (Haynes 2005). The POC fraction consists primarily of plant residues in various stages of decomposition, derived mainly from roots in undisturbed soils (Cambardella and Elliott 1992), and has an important role in the supply of cellular C and energy for microorganisms and in the maintenance of soil structure, particularly macroaggregation (Six et al. 2002).

The conventional tillage system led to a decrease of HWC of about 24% at 0–10 cm depth, a pattern that was already reported in other studies (Chen et al. 2009; Ghani et al. 2003). The negative influence associated with the CT system on HWC concentration has been attributed to the fact that physical disruption of macroagregates promotes the microbial decomposition of this C fraction (Ghani et al. 2003).

Among the studied labile soil organic C fractions, only the AC fraction showed to be sensitive to assess changes in the 0–20 cm (set of two depths) associated with soil management systems. Compared to CT, NT showed significant effects on AC concentration (more 25%) (Table 2). This labile C fraction has been suggested to be a valuable indicator of soil quality under Mediterranean conditions (Melero et al. 2009), and as an indicator of the organic fraction in the environmental monitoring programmes for arid regions, when changes in soil results from slow dynamics processes (Oyonarte et al. 2007). Furthermore, as the determination of AC (using the method of Weil et al. 2003) is a relatively simple method, it could be used in routine soil analyses to differentiate the effects of soil management system on SOM dynamics (Melero et al. 2009).

Besides the observed differences between C fractions concentrations in the first soil layer of CT and NT, a significant decrease was also recorded, with increasing depth mainly in the NT system (Table 3), emphasizing the strong effect of this system on the stratification of labile soil organic C with soil depth. From the overall results of soil organic C fractions, it may suggest that the NT is a suitable system to promote C stock in soil labile fractions of chestnut orchards. This higher amount of C may be linked to the slower decomposition rate as a consequence of stability of aggregates that protect SOM, lower soil temperature, which delay decomposition, and soil erosion decreasing by surface residue maintenance (Bayer et al. 2004). Our results provide additional evidence to previous works (Ramesh et al. 2015), that show the fact that these labile C fractions are useful as indicators for the effect of soil management practices in C dynamics,
Table 3

Pearson correlation between concentration of organic C in mineral soil layers (OC), active C (AC), particulate organic C (POC), hot-water extractable C (HWC) and mineral-associated C (MAC) under no-tillage (NT) and conventional tillage (CT) systems at 0–20 cm soil mineral layer

 

AC

POC

HWC

MAC

NT

 OC

0.805***

0.794***

0.809***

0.907***

 AC

 

0.699***

0.923***

0.689***

 POC

  

0.781***

0.465**

 HWC

   

0.639***

CT

 OC

0.830***

0.736***

0.803***

0.933***

 AC

 

0.798***

0.825***

0.677***

 POC

  

0.878***

0.446*

 HWC

   

0.598***

* Significant at the 0.05 level

** Significant at the 0.01 level

*** Significant at the 0.001 level

Soil mineral-associated C

In the present study, no significant differences were detected in the mineral-associated C (MAC) concentration, when comparing soil management systems (Table 2). The lack of variation caused by different tillage systems on the MAC concentration has already been reported by Freixo et al. (2002) and in other agroforestry systems, may be due to the organic matter stability associated with the interactions with mineral fraction and physical protection by microaggregates (Bayer et al. 2004), leading to detecteble changes only after a considerable time, making MAC not be a good indicator of the effects of land use systems on soil properties (Nogueira et al. 2016). Significant differences were observed, within each management system, being MAC concentration higher in surface (0–10 cm) when compared to the subsurface (10–20 cm) soil layer.

Correlation coefficients, LDA and PCA

The labile soil organic C fractions (AC, POC and HWC) were significantly and positively correlated with each other, with OC and MAC for the set of two depths (0–20 cm) (Table 3), being this type of correlations in agreement with those reported by Chen et al. 2009). The highest r values at α < 0.001 were found in the correlations between OC and MAC, in both systems, AC and HWC, in NT, and POC and HWC in CT (Table 3). Such correlations suggest that OC was a major determinant of the labile soil organic C fractions. Thus, depletion in labile C fractions could give an early indication of the status of SOM, when NT and CT systems are compared. The correlations among the labile soil organic C fractions studied are not surprising since AC, POC and HWC are closely interrelated fractions (Haynes 2005).

Furthermore, the use of a stepwise LDA, which resulted in a discriminant model with two significant discriminant functions that explained 100% of the variance (the first explaining 83.4% and the second 16.6%) (Fig. 1) appears to confirm this statement. Although all C fractions were entered in the LDA analyzes, the model was only based in OC, POC and AC soil contents. Even if the model does not allow a clear separation of different soil management system (NT and CT), it allows the recognition of a pattern, concerning soil C content and depth. A perceptible tendency is visible, separating NT data from CT data: most of the NT data are in the negative region of Factor 2, while most of the CT data are in the positive region of Factor 2. A similar distribution was observed when performing a PCA (Fig. 2), that analysis showed that almost 93% of the total variance of the data could be explained using only two principal components. Using the data of all C fractions, there is a clear separation of the sampling depth (0–10 cm in the positive region of the first factor, and 10–20 cm in the negative region of Factor 1). Furthermore, and as reported for the LDA, a pattern can be observed, concerning soil management system (NT or CT). In fact, most of the samples from CT are located in the positive region of Factor 2, mainly due to the high contents of MAC, while samples from NT are located in the negative region of this same Factor, due to higher amounts of AC, POC and HWC.
Fig. 1

Linear discriminant analysis of the different soil management practices represented in a plane composed by the two main discriminant functions. The functions explain 100% of the total variance. CT conventional tillage, NT no-tillage

Fig. 2

Principal component analysis obtained from different carbon fractions data. 1 CT-0–10 cm; 2 NT-0–10 cm; 3 CT-10–20 cm; 4 NT-10–20 cm. PCA factors explain 92.957% of the total variance

Proportions of organic C fractions in soil organic C

In both management systems, the MAC fraction was the main organic C fraction presented in the soil OC, ranging from about 70% (in the NT, at 0–10 cm depth) to 86% (in the CT, at 10–20 cm depth) (Table 4). The depletion of labile soil organic C fractions in the CT system leads to a redistribution of SOM from labile fractions to the more humified fractions, which may have negative effects on nutrient availability, taking into account that nutrients are mostly dependent on labile fractions (Cambardella and Elliott 1992). Among the labile soil organic C fractions, the POC fraction showed the highest proportion (about 30%) in the NT at surface soil layer, probably related to the decreasing losses of this fraction, and to the increasing promoted by plant residues in the surface soil layer (0–10 cm).
Table 4

Percentage of active C (AC), particulate organic C (POC), hot-water extractable C (HWC) and mineral-associated C (MAC) in organic C (OC) of 0–10 and 10–20 cm soil layers from the no-tillage (NT) and conventional tillage (CT) systems

Depth (cm)

Tillage

AC

POC

HWC

MAC

0–10

NT

2.22

29.85

4.31

70.22

CT

1.75

22.35

3.47

77.59

10–20

NT

1.43

17.25

3.20

82.75

CT

1.27

14.23

3.03

85.77

Conclusions

The obtained results in the present study showed no significant differences in OC concentration between management systems, which may be related to the higher proportion of the recalcitrant C fraction. In contrast, the labile C fractions, AC, POC and HWC showed changes induced by the management system, being significantly higher in the NT as compared to CT, at the soil surface layer (0–10 cm depth). As a consequence, these fractions may be used as sensitive and early indicators for assessing the effects of soil management practices on SOM dynamics and on soil quality. The NT management of chestnut plantations in Mediterranean conditions (maintenance of spontaneous grass cover, controlled by route grazing with sheep and cutting with heavy-duty mulcher in spring and before fruit harvest) seems to be beneficial, in order to enhance soil C sequestration and soil quality.

Notes

Acknowledgements

This study was supported by: European Investment Funds by FEDER/COMPETE/POCI—Operacional Competitiveness and Internacionalization Programme, under Project POCI-01-0145-FEDER-006958 and National Funds by FCT—Portuguese Foundation for Science and Technology, under the Project UID/AGR/04033. The authors thank the staff of the Soil Laboratory of the Instituto Superior de Agronomia (Universidade de Lisboa) for the processing of some of the analyses. Mr. Lindolfo Afonso, landowner of the farm where the orchards are located, is acknowledged for the provided facilities to the study development and José Carlos Rego for assistance in field and laboratory activities.

References

  1. Agroconsultores and COBA (1991) Carta dos Solos, Carta do Uso Actual da Terra e Carta de Aptidão da Terra do Nordeste de Portugal. UTAD/PDRITM, Vila Real, PortugalGoogle Scholar
  2. Alavalapati J, Shrestha R, Stainback G, Matta J (2004) Agroforestry development: an environmental economic perspective. Agrofor Syst 61:299–310Google Scholar
  3. Bayer C, Martin-Neto L, Mielniczuk J, Pavinato A (2004) Armazenamento de carbono em frações lábeis da matéria orgânica de um Latossolo Vermelho sob plantio direto. Pesq Agropec Bras 39:677–683CrossRefGoogle Scholar
  4. Blair G, Lefroy R, Lisle L (1995) Soil carbon fractions based on their degree of oxidation, and the development of a carbon management index for agricultural systems. Aust J Agric Res 46:1459–1466CrossRefGoogle Scholar
  5. Blake G, Hartge K (1986) Bulk density. In: Klute A (ed) Methods of soil analysis, 2nd edn. ASA, Madison, pp 363–375Google Scholar
  6. Cambardella C, Elliott E (1992) Particulate soil organic-matter changes across a grassland cultivation sequence. Soil Sci Soc Am J 56:777–783CrossRefGoogle Scholar
  7. Chen H, Hou R, Gong Y, Li H, Fan M, Kuzyakov Y (2009) Effects of 11 years of conservation tillage on soil organic matter fractions in wheat monoculture in Loess Plateau of China. Soil Tillage Res 106:85–94CrossRefGoogle Scholar
  8. De Leenheer L, Van Hove J (1958) Determination de la teneur en carbone organique des sols. Etude critique des méthodes titrimétriques. Pédologie 8:39–77Google Scholar
  9. Duval M, Galantini J, Iglesias J, Canelo S, Martinez J, Wall L (2013) Analysis of organic fractions as indicators of soil quality under natural and cultivated systems. Soil Tillage Res 131:11–19CrossRefGoogle Scholar
  10. Egnér H, Riehm H, Domingo W (1960) Untersuchungen iber die chemische Bodenanalyse als Grundlag fur die Beurteilunger des Nahrstoffzustandes der Boden. II. Chemische Extractionsmethoden zur Phosphor, und Kalium-bestimmung. Kunglila Lantbrukshögskolans Annaler 26:199–215Google Scholar
  11. Eichhorn M, Paris P, Herzog F, Incoll L, Liagre F, Mantzanas K, Mayus M, Moreno G, Papanastasis V, Pilbeam D, Pisanelli A, Dupraz C (2006) Silvoarable systems in Europe—past, present and future prospects. Agrofor Syst 67:29–50CrossRefGoogle Scholar
  12. Freixo A, Machado P, Guimarães C, Silva C, Fadigas F (2002) Estoques de carbono e nitrogênio e distribuição de frações orgânicas de Latossolo do Cerrado sob diferentes sistemas de cultivo. R Bras Ci Solo 26:425–434CrossRefGoogle Scholar
  13. Ghani A, Dexter M, Perrott K (2003) Hot-water extractable carbon in soils: a sensitive measurement for determining impacts of fertilization, grazing and cultivation. Soil Biol Biochem 35:1231–1243CrossRefGoogle Scholar
  14. Haynes R (1980) Influence of soil management practice on the orchard agroecosystem. Agro-ecosystems 6:3–32CrossRefGoogle Scholar
  15. Haynes R (2005) Labile organic matter fractions as central components of the quality of agricultural soils: an overview. Adv Agron 85:221–268CrossRefGoogle Scholar
  16. Hernández A, Lacasta C, Pastor J (2005) Effects of different management practices on soil conservation and soil water in a rainfed olive orchard. Agric Water Manag 77:232–248CrossRefGoogle Scholar
  17. Houba V, Van Der Lee J, Novozamsky I, Walinga I (1986) Soil and plant analysis procedures. Wageningen University, WageningenGoogle Scholar
  18. IPMA (2013) Normais Climatológicos 1971–2000 de Bragança. Instituto de Meteorologia, IP Portugal. http://www.meteo.pt/pt/oclima/clima.normais/003/. Accessed 11 Mar 2013
  19. Jose S (2009) Agroforestry for ecosystem services and environmental benefits: an overview. Agrofor Syst 76:1–10CrossRefGoogle Scholar
  20. Lal R (2003) Offsetting global CO2 emissions by restoration of degraded soils and intensification of world agriculture and forestry. Land Degrad Dev 14:309–322CrossRefGoogle Scholar
  21. Lalitha M, Kumar P (2016) Soil carbon fractions influenced by temperature sensitivity and land use management. Agrofor Syst 90:961–964CrossRefGoogle Scholar
  22. Maroco J (2003) Análise Estatística, com utilização do SPSS. Edições Sílabo, LisboaGoogle Scholar
  23. Martins A, Raimundo F, Borges O, Linhares I, Sousa V, Coutinho J, Gomes- Laranjo J, Madeira M (2010) Effects of soil management practices and irrigation on plant water relations and productivity of chestnut stands under Mediterranean conditions. Plant Soil 327:57–70CrossRefGoogle Scholar
  24. Martins A, Marques G, Borges O, Portela E, Lousada J, Raimundo F, Madeira M (2011) Management of chestnut plantations for a multifunctional land use under Mediterranean conditions: effects on productivity and sustainability. Agrofor Syst 81:175–189CrossRefGoogle Scholar
  25. Melero S, López-Garrido R, Murillo J, Moreno F (2009) Conservation tillage: short-and long-term effects on soil carbon fractions and enzymatic activities under Mediterranean conditions. Soil Tillage Res 104:292–298CrossRefGoogle Scholar
  26. Moreno B, Garcia-Rodriguez S, Cañizares R, Castro J, Benitez E (2009) Rainfed olive farming in south-eastern Spain: long-term effect of soil management on biological indicators of soil quality. Agric Ecosyst Environ 131:333–339CrossRefGoogle Scholar
  27. Nair P, Nair V, Kumar B, Showalter J (2010) Carbon sequestration in agroforestry systems. Adv Agron 108:237–307CrossRefGoogle Scholar
  28. Nogueira L, Silva D, Pereira M, Gaia-Gomes J, Silva E (2016) Biological properties and organic matter dynamics of soil in pasture and natural regeneration areas in the Atlantic forest biome. Rev Bras Cienc Solo 40:e0150366CrossRefGoogle Scholar
  29. Nyamadzawo G, Nyamangara J, Nyamugafata P, Muzulu A (2009) Soil microbial biomass and mineralization of aggregate protected carbon in fallow-maize systems under conventional and no-tillage in Central Zimbabwe. Soil Tillage Res 102:151–157CrossRefGoogle Scholar
  30. Oyonarte C, Mingorance M, Durante P, Piñero G, Barahona E (2007) Indicators of change in the organic matter in arid soils. Sci Total Environ 378:133–137CrossRefPubMedGoogle Scholar
  31. Paudel B, Udawatta R, Anderson S (2011) Agroforestry and grass buffer effects on soil quality parameters for grazed pasture ad row-crop systems. Appl Soil Ecol 48:125–132CrossRefGoogle Scholar
  32. Raimundo F (2003) Sistemas de mobilização do solo em soutos. Influência na produtividade de castanha e nas características físicas e químicas do solo. PhD dissertation. Universidade de Trás-os-Montes e Alto Douro, Vila Real, PortugalGoogle Scholar
  33. Ramesh T, Manjaiah K, Mohopatra K, Rajasekar K, Ngachan S (2015) Assessment of soil organic carbon stocks and fractions under different agroforestry systems in subtropical hill agroecosystems of north-east India. Agrofor Syst 89:677–690CrossRefGoogle Scholar
  34. Six J, Feller C, Denef K, Ogle S, Sá J, Albrecht A (2002) Soil organic matter, biota and agreggation in temperate and tropical soils; effects of no-tillage. Agronomie 22:755–775CrossRefGoogle Scholar
  35. Stamps W, Linit M (1999) The problem of experimental design in temperate agroforestry. Agrofor Syst 76:187–196Google Scholar
  36. Thomas G (1982) Exchangeable cations. In: Page A, Miller R, Keeney D (eds) Methods of soil analysis. Part 2 agron 9, 2nd edn. ASA, SSSA, Madison, pp 159–165Google Scholar
  37. Weerasekara C, Udawatta R, Jose S, Kremer R, Weerasekara C (2016) Soil quality differences in a row-crop watershed with agroforestry and grass buffers. Agrofor Syst 90:829–838CrossRefGoogle Scholar
  38. Weil R, Islam K, Stine M, Gruver J, Samson-Liebig S (2003) Estimating active carbon for soil quality assessment: a simplified method for laboratory and field use. Am J Altern Agric 18:3–17CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media Dordrecht 2017

Authors and Affiliations

  1. 1.Direção Regional de Agricultura e Pescas do NorteDelegação do Nordeste TransmontanoBragançaPortugal
  2. 2.Centre for the Research and Technology of Agro-Environmental and Biological Sciences - CITABUniversity of Trás-os-Montes e Alto Douro - UTADVila RealPortugal
  3. 3.Department of Biology and EnvironmentUniversity of Trás-os-Montes and Alto DouroVila RealPortugal
  4. 4.Centro de Estudos Florestais, Instituto Superior de AgronomiaUniversidade de LisboaLisbonPortugal

Personalised recommendations