3.1 CO2 fluxes
The mean variation of soil CO2 flux (CO2-F) for the residue management systems during the experimental period ranged from 0.7 to 1.3 kg ha−1 h−1 (Figure 6-A1). The results are consistent with CO2-F values found in other studies carried out in commercial eucalyptus forests in Brazil: 1.19 - 1.82 kg ha−1 h−1 [34]; 1.31 - 1.56 kg ha−1 h−1[35]; 1.28 kg ha−1 h−1 [36]; 0.75 - 1.51 kg ha−1 h−1 [37].
The higher CO2 fluxes (CO2-F) in areas with HR+L, observed at 0.3 and 9.3 months after harvest (MAH) (Figure 6-A1), are due to the greater amounts of residues present in this treatment, which increases the rate of microbial activity. In a similar study, [14] found average CO2-F of 0.93 kg ha−1 h−1 in an area without residues and 1.50 kg ha−1 h−1 in an area where eucalyptus harvest residues and litter from the previous rotation were left. The increase in the population of soil microbial biomass is associated with the amounts of decomposable residues left in the area [38, 12], which must have occurred in the period corresponding to 15.8 and 22.3 MAH, leading to a reduction in CO2-F (Figure 7A1). Similar results were observed by [39], who found that, immediately after sugarcane harvest, CO2 emissions increased linearly with the increase in the amount of straw deposited on the soil, followed by a decrease six months after the deposition of plant residues on the soil.
In the area where all residues were removed (WR), at 9.3 MAH, CO2-F was equal to 1.11 kg ha−1 h−1, being higher than that of the L treatment and similar to that of the HR+L treatment. The higher fluxes may also have been stimulated by higher soil temperature (Ts) (Table 3). The sampling performed at 9.3 MAH coincided with the dry period, when high temperatures are recorded (Figure 1). Additionally, in this sampling period, the soil was more exposed to solar radiation, considering the low size of the plants and, consequently, less shading. The uncovered soil, without the presence of residues, is more exposed to solar radiation, which in turn increases Ts [40]. Increments in Ts stimulate microbial respiration, which in turn increases CO2 emissions to the atmosphere [41, 42]. [43] observed that there was higher CO2-F in soils that were uncovered and without plant residues due to higher Ts values.
In the treatment where only litter from the previous rotation was left (L), the CO2-F remained constant as a function of the collection times (Figure 6-A1). Litter has a greater amount of materials with higher degree of recalcitration, in which most of the labile C has already suffered interference from decomposer microbial activity [44]. C:N ratio in the L treatment was 42% higher than in the HR+L treatment. In addition, the quantity of C is three times lower in L (Table 2).
Throughout the experiment, regardless of the treatments, higher (p < 0.05) CO2-F values were observed in the planting rows compared to the inter-rows, even at 9.3 MAH (Figure 6-A2). Eucalyptus planting rows have higher CO2-F than the inter-rows [11, 23]. [45] noted a gradual increase in GHG emissions as the samplings approached the base of the pine tree.
Planting rows are sites with higher root volume, due to the proximity of the stumps; in addition, the growth of the trees is favored by the better physical conditions of the soil promoted by minimum tillage [46] particularly in cohesive soils such as that of the present experiment [47]. Moreover, fertilization in eucalyptus plantations is directed to the planting row region [48]. Greater amounts of roots increase CO2-F as a consequence of root respiration [49]. The higher CO2-F levels found in the planting rows at 0.3 and 9.3 MAH may result, although partially, from the death and renewal of the roots of sprouts. [50] suggest that, up to 60 days after cutting eucalyptus trees, there is high mortality of thin roots in the 0.0-1.0 m layer, which contributes to C supply along the soil profile. The rhizosphere is a site enriched with exudates, mucilage, lysates and secretions that can be used as substrate by the various groups of soil microorganisms, in interactions developed with the root system [51].
Another factor that may explain the higher CO2-F observed in the planting rows compared to the inter-rows is the lower bulk density (Ds) found in this position in all soil layers evaluated (Figure 3). High Ds, and the consequent decrease in soil porous space, limits the diffusion of gases [52, 53] After harvest, there was a significant increase in Ds in the 0-10 and 60-100 cm layers in the inter-rows (Figure 4). The same did not occur in the planting rows, since machine traffic occurs predominantly in the inter-rows [54].
No significant correlation was observed between CO2-F and soil moisture and temperature (Table 4). The results of the present study differ from those reported by [2, 53, 55] Neto et al. (2011), Bicalho et al. (2017) and Vicentini et al. (2019), who found significant correlations between soil moisture and temperature and CO2-F. However, these studies evaluated the influence of soil moisture and temperature as well as CO2-F separately, without the presence of organic residues. This is evidence that, in this case, the decomposition of organic residues in the soil also influences carbon dioxide emissions, as well as soil moisture and temperature. [14] conducting an experiment in tropical environment, as in the present study, found no influence of eucalyptus harvest residues on soil moisture and temperature in most of the experimental time (2 years).
3.2 CH4 fluxes
The observed CH4 fluxes ranged from 0.001 to -0.0006 kg ha−1 h−1 (Figure 6C1), with magnitude similar to that found in other experiments in commercial eucalyptus forests: -0.00002 to -0.00015 kg ha−1 h−1 [57]; 0.0027 to -0.0028 kg ha−1 h−1 [58]; -0.00010 to -0.0009 kg ha−1 h−1 [59]; 0.0001 to -0.0006 kg ha−1 h−1 [60]; -0.0002 to -0.0007 kg ha−1 h−1 [61].
In this study, the CH4 fluxes to the atmosphere (CH4-F), or influxes to the soil (CH4-I), were temporarily related to the presence of plant residues on the soil (Figure 6-B1). The highest CH4-F values (0.0009 kg ha−1 h−1) were found in the pre-harvest. After harvest, fluxes were observed in the collections performed at 0.3 and 9.3 MAH, in treatments with litter (L) or harvest residues plus litter (HR+L). Influxes were observed in these same treatments in the collections performed at 15.8 and 22.3 MAH. In the sites with no residues (WR), there was a reduction in CH4-F compared to the pre-harvest, and influxes were observed at all sampling times.
At five days before harvest (PH) and at 0.3 MAH, collections that coincided with the rainy season, the large amount of residues on the soil in the treatments (L and HR+L) and PH may have reduced the diffusion of gases between the atmosphere and the soil. [62] in a study conducted in forest areas, found a 16% increase in CH4 influx with the removal of litter. This result was attributed to the fact that litter hinders the diffusion of gases [63]. In addition, this situation of higher soil moisture (Us) and presence of residues may favor higher microbial respiration, increasing sites with low concentrations of O2 [64]. The Us values at these evaluation times were above field capacity in the sampled layers (Table 2 and 3). Characteristics of the soil, that is, presence of textural B horizon, of cohesive nature, and flat topography may have contributed to a lower water drainage in the soil profile [65] and consequent increase in CH4 flux. Rainfall events with greater volumes promote environments with low amount of O2, stimulating the activity of methanogenic microorganisms, which decompose the organic matter available in the soil, producing CH4 [38].
and consequent increase in CH4 flux. Rainfall events with greater volumes promote environments with low amount of O2, stimulating the activity of methanogenic microorganisms, which decompose the organic matter available in the soil, producing CH4 [66, 67] During the decomposition of forest residues, there may be emission of CH4 caused by fungi [68] It has been observed in some studies that UV radiation and high temperature on dry and fresh leaves and on structural components, including pectin, lignin and cellulose, lead to the increase of CH4-F [69, 70, 71].
At 15.8 and 22.3 MAH, CH4 emissions were reduced in the treatments L and HR+L (Figure 6-B1). The reduction in the amount of residues, especially leaves, and the consequent reduction in C contents may have led to reduction in CH4 emissions (Figure 7). [39] described similar results when detecting CH4 emissions in areas with sugarcane straw up to six months after harvest, with a reduction after that, due to the advanced degree of decomposition of the residues.
In the collections performed at 0.3 and 9.3 MAH in the HR+L treatment, CH4 fluxes were more pronounced in the planting rows. This position and these sampling times also had higher CO2 fluxes, which possibly contributed to the generation of anaerobiosis sites. Unavailability of O2 is the main factor causing CH4 emission from the soil to the atmosphere [72].
There was no significant correlation between Us and Ts and CH4 fluxes or influxes in rows and inter-rows (Table 4). In the treatment L or HR+L, the possible hypoxia events in the rainy season and the release of CH4 by the action of temperature and UV radiation in the dry season may have been equated. In addition, at 15.8 and 22.3 MAH, CH4 fluxes were observed in all treatments, regardless of the evaluation time (dry or rainy season) (Figure 6-B1).
3.3 N2O fluxes
N2O flux ranged from 0.0009 kg ha−1 h−1 to -0.0009 kg ha−1 h−1 (Figure 6C1), corroborating results reported in previous experiments in commercial eucalyptus forests, whose variations were: 0.0011 to -0.00037 kg ha−1 h−1[58]; 0.00015 to -0.00005 kg ha−1 h−1 [73]; 0.00015 to -0.00005 [74] and 0.00098 kg ha−1 h−1 [75].
N2O fluxes (N2O-F) to the atmosphere, or influxes to the soil (N2O-I), varied with the different residue managements (Figure 6-C1). In the pre-harvest (PH) and HR+L treatment, the highest N2O-F values were observed at 0.3 and 9.3 MAH. In comparison with PH, the total (WR) or partial (L) removal of residues caused reduction in N2O-F to the atmosphere. In the L treatment, fluxes were observed only at 0.3 MAH. In the WR treatment, influxes were observed at all sampling times.
Residue management is among the main factors that influence nutrient cycling in eucalyptus plantations [30]. Nitrogen in organic forms, such as the one added to the soil in the form of residues, can be converted into N2O [76]. With the partial or total removal of residues, there is a reduction in N stocks (Table 2), limiting N availability in the soil, which leads to reduction in N2O emission [77].
In this experiment, for the situation corresponding to HR+L, where there was a greater amount of residues on the soil, higher N2O-F values were observed at 0.3 and 9.3 MAH (Figure 6-C1). When litter and the residues generated in eucalyptus harvesting are kept in the area of the stand, N content in the soil increases [78]. Results obtained by [79, 58] demonstrate the relationship between the amount of residues on the soil and N2O-F. [80] observed that the N2O-F of an area is closely related to the N released from the residues maintained in that area.
Also in HR+L, the evaluations performed at 15.8 and 22.3 MAH showed N2O influxes (Figure 6-C1). There was a significant release of N up to 10 months; after that, the quantities and the rate of release of the nutrient contained in the residues decreased (Figure 7A2). The leaves present in the residues have high lability for decomposition, so there was accelerated decomposition of this organ and great release of N in the first months (Figure 7A1). Similar results were observed by [81].
In the L treatment, N2O fluxes were observed at 0.3 MAH (Figure 6-C1). However, the emission of this gas was lower than in the HR+L treatment. In the beginning of the experiment, the quantity of N in the L treatment was approximately 450% lower than in the HR+L treatment (Table 2). N2O influxes were observed in the samplings performed at 9.3, 15.8 and 22.3 MAH (Figure 6-C1). In addition to the low quantities of N in the residues (Table 2), there was also a lower N release speed compared to the HR+L treatment (Figure 7B2). Litter is a material with high degree of decomposition. Compared to the HR+L treatment, the C:N ratio of litter was 42% higher than that found under the condition in which it was combined with the other residues, which led to a longer half-life time (Table 2).
There was no significant correlation between Us, Ts and N2O fluxes or influxes (Table 4). In the treatments HR+L and L, N2O-F occurred in the rainy season; the presence of residues may have reduced the diffusion of gases at the atmosphere-soil interface. Soil moisture is the variable that most favors N2O-F to the atmosphere [82] With high Us, certain microorganisms can use NO3− as final electron acceptors in place of O2 [83].
The higher N2O-F values observed in the HR+L treatment in the dry period may be linked to the high temperatures in the soil (Table 3). Temperature influences N2O emissions from the soil to the atmosphere, and denitrification can be extremely sensitive to rising temperatures, as increased microbial respiration results in O2 depletion [84, 85]. In the present study, when high CO2 fluxes were observed (HR+L treatment; at 0.3 and 9.3 MAH - Figure 6-A1), higher N2O fluxes were also observed (Figure 6-C1) compared to the other treatments. According to [86] the addition of plant residues on the soil rapidly stimulates microbial activity, creating anaerobiosis sites and, consequently, stimulating the denitrification process.
Table 3
Soil temperature (Ts) and soil moisture (Us) measured at 5 cm depth, in the row and inter-row of eucalyptus plantation, 5 days before harvest (pre-harvest) and at 0.3, 9.3, 15.8 and 22.3 months after harvest. The samples were collected in the following treatments: without the presence of residues (WR), with litter from the previous rotation (L) and with harvest residues and litter from the previous rotation (HR+L)
Sampling season
|
Treatment
|
Planting row
|
Planting inter-row
|
|
|
Ts
|
Us
|
Ts
|
Us
|
|
|
(°C)
|
(%, v v−1)
|
(°C)
|
(%, v v−1)
|
|
Pre-harvest
|
Rainy period
|
-
|
29.7
|
17.4
|
31.3
|
16.6
|
|
0.3 Months after harvest
|
Rainy period
|
WR
|
27.0 a
|
19.8 a
|
28.5 a
|
17.6 a
|
L
|
27.0 a
|
17.5 a
|
27.5 ab
|
16.8 a
|
HR+L
|
27.0 a
|
17.3 a
|
27.0 b
|
16.0 a
|
|
CV (%)
|
0
|
15.7
|
1.8
|
10.3
|
|
9.3 Months after harvest
|
Dry period
|
WR
|
29.5 a
|
1.8 a
|
31.0 a
|
1.5 a
|
L
|
29.0 ab
|
2.3 a
|
30.5 ab
|
3.0 a
|
HR+L
|
28.25 b
|
2.1 a
|
29.0 b
|
2.1 a
|
|
CV (%)
|
1.7
|
35.9
|
3.2
|
43.1
|
|
15.8 Months after harvest
|
Rainy period
|
WR
|
22.5 a
|
22.2 a
|
22.5 a
|
23.9 a
|
L
|
22.5 a
|
24.2 a
|
22.1 a
|
22.0 a
|
HR+L
|
22.5 a
|
23.8 a
|
22.7 a
|
25.6 a
|
|
CV (%)
|
1.3
|
5.3
|
1.7
|
12.8
|
|
22.3 Months after harvest
|
Dry period
|
WR
|
25.3 a
|
8.1 b
|
24.7 a
|
8.4 b
|
L
|
25.5 a
|
8.8 b
|
25.4 a
|
10.2 a
|
HR+L
|
25.3 a
|
10.5 a
|
25.4 a
|
10.6 a
|
|
CV (%)
|
2.1
|
7.8
|
1.3
|
8.5
|
Means followed by equal lowercase letters vertically do not differ from each other by Tukey test (p < 0.05). CV (%) denotes the coefficient of variation around the mean (n=8).
Table 4
Pearson’s correlation coefficients between CO2, CH4 and N2O fluxes and soil temperature and moisture measured in the planting row (R) and inter-row (IR) in areas without the presence of residues (WR), with litter from the previous rotation (L) and with harvest residues and litter from the previous rotation (HR+L)
|
Greenhouse gases
|
Description
|
CO2
|
CH4
|
N2O
|
Sampling position
|
R
|
IR
|
R
|
IR
|
R
|
IR
|
Soil moisture
|
-0.24 ns
|
0.01ns
|
-0.18 ns
|
-0.27 ns
|
-0.10 ns
|
-0.06 ns
|
Soil temperature
|
0.22 ns
|
0.12 ns
|
0.05 ns
|
0.54 ns
|
0.16 ns
|
0.09 ns
|
ns Non-significant correlation; * Significant correlation (p < 0.05) |
3.4 CO2 equivalent flux
CO2eq-F was similar at all evaluation times when comparing the WR and L managements. In the HR+L treatment, the flux was higher than in the others, at 0.3 and 9.3 MAH. In the other samplings, the fluxes were similar to those found in the other treatments (Figure 8-A). The high C and N release rates in the HR+L treatment in the first months, via decomposition (Figure 7-A2), can be explained by the high amount of residues (34.7 t ha−1).
During the decomposition process, plant residues release mostly C and N, of which, on average, 50 and 70%, respectively, return to the atmosphere in oxidized forms [20]. Despite the higher CO2eq-F in the HR+L treatment, it should be noted that with the maintenance of residues, C and N inputs to the soil are expected [19, 87, 12] which could increase C and N stocks in the soil, constituting an important strategy for reducing GHG concentrations in the atmosphere in forestry and agricultural systems [88].
When analyzing the contributions of gases to the overall emissions, a quantitatively high contribution of CO2 is observed (Figure 6-A1). CH4 and N2O emissions or influxes were similar (Figure <link rid="fig6">6</link>-B1 and 6-C1), but N2O has a higher warming potential compared to CH4 [25]. In the HR+L treatment, the highest CO2 fluxes were observed in the collections performed at 0.3 and 9.3 MAH (Figure 6-A1). Higher CH4 and N2O fluxes were also observed (Figure <link rid="fig6">6</link>-B1 and 6-C1). The close relationship between CO2 fluxes in the soil and CH4 and N2O fluxes is reported in other studies. [64] conducting an experiment under controlled conditions, artificially injected CO2 to 10 cm deep into the soil through a probe and observed that N2O and CH4 emissions increased linearly with CO2 concentration and time of application.
3.5. Quantities of carbon and nitrogen in plant biomass and soil
The tree biomass and its quantity of C at 22.3 MAH did not differ under the influence of residue management (Table 5), despite the nutrients released by eucalyptus harvest residues. Positive responses in the growth of eucalyptus plants in the presence of residues in soils with low levels of organic matter, P, K, Ca, Mg, and S are attributed mainly to the release of nutrients from the residues to the soil [14, 89]. In this experiment, the absence of response to the maintenance of residues may be due to the existence of nutrients at appropriate levels to the plants (Table 1), both natural from the soil and residual from the fertilizers and limestone applied in previous rotations. Similar results have been reported by other authors [90]. Increases in forest productivity only after two successive rotations of harvest residue maintenance in the same area [12].
Table 5
Quantity of carbon in tree biomass and soil, at 22.3 months after eucalyptus harvest, in the area without the presence of residues (WR), with litter from the previous rotation (L) and with harvest residues and litter from the previous rotation (HR+L). Means followed by equal lowercase letters vertically do not differ from each other by Tukey test (p < 0.05). CV (%) denotes the coefficient of variation around the mean (n=4)
|
Plant biomass
-------------------------------
|
Quantity of carbon
----------------------------------
|
Treatments
|
Below ground1
|
Above ground
|
Total
|
Below ground1
|
Above ground
|
Total
|
|
------------------------- Mg ha−1 --------------------------
|
WR
|
0.67 a
|
63.92 a
|
64.57 a
|
0.28 a
|
28.86 a
|
29.13 a
|
L
|
1.18 a
|
62.76 a
|
63.94 a
|
0.52 a
|
26.71 a
|
27.22 a
|
HR+L
|
0.92 a
|
65.34 a
|
66.26 a
|
0.41 a
|
27.57 a
|
27.99 a
|
CV (%)
|
36.44
|
8.42
|
8.59
|
42.83
|
8.43
|
8.68
|
1 0-40 cm soil layer |
The removal of residues from planted forests aimed at bioenergy production, for instance, leads to the export of soil nutrients [15, 77, 91], hence requiring greater nutritional replacements via fertilizers. In such situations, there may be greater release of GHG to the atmosphere, resulting from the manufacture of fertilizers and their contact with the soil [92, 93, 94].
C (Figure 10A) and N (Figure 10B) stocks in the soil surface layer (0-10 cm) at 22.3 MAH were similar in all treatments, despite the different quantities of C and N supplied to the area in the form of residues (Table 2). Changes in soil C stocks are slow, with difficult detection in the laboratory in short experimental periods [95]. Therefore, long-term studies for monitoring C stocks in soil are essential to determine these changes, which are often noticeable only after several years or decades [96].
The absence of increase of soil organic matter in the HR+L and L treatments compared to the treatment without residues (WR) may also be due to the fact that the experimental area is in the third rotation and received residues from the harvest of two previous rotations successively. [14] also found no increase in organic matter content in the soil with retention of eucalyptus harvest residues. These authors attributed this observation to the low quality of the residues. [97] in eucalyptus forests, observed that regardless of the quantity and quality of litter on the soil surface there was no influence on soil organic C contents. The input of C in quantity and quality in forest soils does not always result in an increase of C in the soil due to the differences in the biochemical characteristics of the residues, mineralogy through the physical protection of SOM and the priming effect in the soil [98].
Additionally, the surface layers of the soil under study are sandy (Table 1), hence offering little physical protection to residues, facilitating microbial decomposition and hindering the formation of stable organic matter [99]. Results similar to these were found by [100] in sandy soil in the third rotation of eucalyptus planting.
3.6. Balance between CO2 equivalent emissions and quantity of C in the plant-soil system
In all treatments, there were higher quantities of C immobilized in the plant-soil system compared to atmospheric GHG emissions, thus indicating C sequestration (Figure 11). There is consensus that planted eucalyptus forests, mainly in tropical or subtropical areas, contribute to greater C sequestration than other forests, mainly due to the accelerated growth of tree biomass [22, 101, 102, 103, 104].
Where the harvest residues were kept (HR+L), C balance was 14.4 and 16.7% lower compared to the treatments in which the residues were partially or totally removed (L and WR), respectively (Figure 11). The lower C balance in the HR+L treatment is due to the higher CO2 equivalent fluxes observed at 0.3 and 9.3 months of experiment, compared to the other treatments (Figure 8). Furthermore, during the experimental period, no differences were observed between treatments in C fixation in plant biomass (Figure 9) and in soil (Figure 10).
Although the CO2eq-F values found in the treatment are higher where harvest residues were left (HR+L), at 0.3 and 9.3 MAH, it should be considered that the partial (L) or total (WR) removal of forest residues will cause GHG emissions. GHG emissions will occur in the process of removing material from the field, during transportation and in the processing in the industry for energy production. In this scenario, the consumption of fossil fuels should be accounted for, considering the removal of residues from the field and transport to industry. Diesel consumption is the main source of GHG emissions throughout the eucalyptus production cycle, corresponding to approximately 64% of the total emission according to [105]. The relevant GHG emissions during biomass combustion for energy generation should also be considered [106].
Additionally, it should be considered that the permanence of forest residues in the planting area favors nutrient cycling [12, 15], contributes to increasing soil diversity and microbial activity [12, 13], attenuates the physical deterioration of the soil under machine traffic [18, 19], reduces soil losses via water erosion [17], reduces thermal oscillations of the soil [107, 108] and minimizes soil water loss via evaporation [108].