CO₂ Flux Emissions by Fixed and Mobile Soil Collars Under Different Pasture Management Practices

Abstract

Carbon dioxide flux emissions (CFE) from agricultural areas exhibit spatial and temporal variability, and the best time of collar fixation to the soil prior to the collection of CO₂ flux, or even its existence as a factor, is unclear. The objective of this study was to evaluate the effect of the fixation time of collars that support the soil-gas flux chamber based on the influence of CFE on different pasture management practices: control (traditional pasture management practice) (CON), chisel (CHI), fertilized (FER), burned (BUR), integrated crop-livestock (iCL), and plowing and harrowing (PH). A field study was conducted on the clayey soil of Udults. The evaluations were performed monthly by fixing the PVC collars 30 d and 30 min prior to each CFE measurement. Although a linear trend in CFE was observed within each pasture management practice between the two collar-fixation times, collar fixation performed 30 min prior led to an overestimation of CFE by approximately 32.7% compared with 30 d of collar fixation. Thus, CFE were higher (p ≤ 0.10) in the MC, when compared to the FC, when the CON, BUR, and iCL managements were evaluated. Overall, fixing the collar 30 d prior to field data collection can improve the quality of the data, making the results more representative of actual field conditions.

1. Introduction

In the contemporary global context, it is recognized that encouraging low-carbon agriculture is essential for mitigating climate change and ensuring food security [1]. High levels of emissions are often caused by anthropogenic activities, particularly those carried out in the agricultural sector [2]. According to the Greenhouse Gas Emissions and Removal Estimation System SEEG, Brazilian agriculture was responsible for up to 75% of national carbon dioxide (CO₂) emissions in 2022 and is the largest source of anthropogenic greenhouse gas emissions worldwide [3]. However, considering the fundamental importance of agriculture in the global economy, which ensures food security and human prosperity [4], the need to reduce carbon emissions has emerged as a global duty and a collective effort [2]. This is a challenging proposition when aiming to balance agricultural production with sustainable, low-carbon development [5]. To this end, efforts must be made in the sectors that make up the production chain. In this context, with the aim of contributing to the reduction of high emissions, several studies with soil management practices and cultural systems that minimize CO₂ flux emissions or maximize soil carbon (C) sequestration have been performed [6,7,8], aiming at the sustainability of agricultural activities. However, the complexity and uncertainties involved in measuring soil CO₂ flux emissions, along with the lack of standardized data collection in several of these studies, often compromise the accuracy of estimates. This, in turn, makes it difficult to compare and extrapolate the results across different regions. Notably, CO₂ flux emissions from agricultural areas exhibit great spatial [9] and temporal variability [10], which are influenced by several attributes of the soil [9,11], as well as by sampling time, climate [12], and soil management practices [13]. The mitigation of uncertainties requires improvement for accuracy in estimating the evaluation of C flow between the atmosphere and terrestrial ecosystems [14].

Although various static chamber models are currently used in GGE inventories in Brazil and worldwide [15,16,17,18,19], shortages in the availability and use of simpler and more economical equipment still exist. The time of collar fixation to the soil prior to collecting CO₂ flux when using infrared gas analyzer devices to measure these emissions requires further study. Although several studies have been carried out testing the size of the collar [20], the insertion depth [21,22,23], and even different types of equipment [19], most do not clarify this time or even address it [15,18,24]. Typically, this fixation occurs at the time of field measurement or a few hours prior [18]. In some situations, the equipment is fixed to the ground for an extended period or permanently [25], but this does not prevent distortions in the results [26].

Even when using manuals of the equipment that use soil collars as support for chambers, the ideal time to wait after the fixation to initiate the CO₂ flux measurement is not evident. Moreover, it is recommended that emissions be measured after installing the collar in the soil until the flow of CO₂ stabilizes LI-8100-104 (LI-COR Biosciences, Lincoln, NE, USA), but this practice can be time-consuming. Given that biased measurements can result from variations in the protocols related to the collar [20], the standardization of the time of collar fixation to the soil prior to CO₂ flux collection can facilitate the planning of data collection in the field, avoid overestimation or underestimation of results, and allow comparisons between similar studies to be more faithful to the processes studied.

Therefore, developing a protocol that standardizes the initial part of data collection is necessary. The objective of the present study was to evaluate the effect of collar fixation time on the soil (30 d and 30 min prior to data collection) and its influence on determining CO₂ flux emissions in different pasture management practices. Thus, this work seeks to fill an important gap in practical research regarding the investigation of CO₂ flux emissions, and to assist in the enhancement of the implementation and monitoring of low-carbon agriculture. This research moves towards standardizing protocols for future studies of carbon emissions in agriculture.

2. Materials and Methods

2.1. Study Area

This study was conducted on an experimental farm at the Center of the Agricultural Science Faculty of the Universidade Federal do Espírito Santo (UFES), located in the municipality of Alegre, Espírito Santo, Brazil. The experimental field is located at 20°74′ S latitude and 41°48′ W longitude (Figure 1). The soil was classified as Udults clayey [27], and the mean altitude of the experimental farm was 155 m, with a slope of 32%. The regional climate is classified as tropical (Aw) with dry winters. The average annual rainfall is 1346 mm, with most of the precipitation occurring between November and March. Data from the meteorological station at the UFES indicate an average annual temperature of 22.2 °C, with maximum and minimum averages of 29 °C and 16.9 °C, respectively. The bulk density of the soil in the area is 1.44 g/cm³, and the soil composition comprises 40.39% clay, 7.43% silt, and 52.17% sand [28].

2.2. Management Characterization

Field experiments were conducted in March and August 2014. The soil pasture management practices were as follows:

Control pasture (CON): This control plot represented pastures managed in the landscape for >10 years. Pasture management practice did not involve the use of lime or fertilizers. The pasture was in the initial stage of degradation, with partially exposed soil at some points. The grazing simulation was performed when the grass reached 20 cm (±5 cm) in height, thus lowering the extract to 10 cm. The same grazing simulation practice was adopted for all other pasture management practices to isolate the effects of differentiated grazing between areas and the difficulty of working with animals in the experimental area.

Chisel pasture (CHI): This included soil chisel management with manual rows in contours. Lime (1 t ha⁻¹) and NPK fertilizers (30 kg ha⁻¹ of K₂O and 110 kg ha⁻¹ of P₂O₅) were used. Brachiria brizhanta seeds were then replanted. After grass emergence, 100 kg ha⁻¹ of N₂O was applied.

Fertilized Pasture (FER): Liming (1 t ha⁻¹) and NPK fertilizer (200 kg ha⁻¹ of K₂O and 50 kg ha⁻¹ of P₂O₅) were applied. After grass emergence, during the rainy season a split application of N₂O was performed using three equal doses of 50 kg ha⁻¹ each (a total of 150 kg ha⁻¹ of N₂O).

Burned pasture (BUR): Fast burning of all residues, including original Brachiaria plants, invasive plants, and litter was performed. The fire was generated using a flamethrower attached to a gas canister. No lime or fertilizer was used in this management.

Integrated crop-livestock management (iCL): Initially residues of Brachiaria brizantha and invasive plants were chemically desiccated with glyphosate^®. Liming (1 t ha⁻¹) was performed 12 d after herbicide application. Manual row planting was implemented in the contour and NPK fertilizers (80 kg ha⁻¹ of K₂O and 120 kg ha⁻¹ of P₂O₅) were applied [29]. Legumes Feijão de Porco (Canavalia ensiformis) and Feijão guandú (Cajanus cajan) were planted and used as cover plants. The legumes were planted by adopting intercalary and manual distribution of seeds along the rows leaving a gap of 0.10 m between each plant. After 60% of the legume plants had flowered (118 d), all were cut, and the residues were maintained as soil cover.

Plowing and harrowing in contour (PH): In the plowing and harrowing in contour simulation, the residues were incorporated into the soil to a 0.15 m depth. No lime or fertilizer was applied.

2.3. Data Collection and Soil Sampling

Carbon dioxide (CO₂) flux emissions were measured within each management system for 6 months, consecutively (March/2014; April/2014; May/2014; June/2014; July/2014; August/2014). The equipment used for the CO₂ flux was LI-8100-104 (LI-COR Biosciences, Lincoln, NE, USA), and the collars were made of PVC pipes, according to the specifications of the LI-8100-104 manual.

The spacing between the mobile collar (MC) and fixed collar (FC) was 10 cm, to ensure that the collection points were close, avoiding soil flux variation due to the collection site but not due to the treatment. Collar fixation was standardized at 5 cm above and below the ground.

Soil temperature was measured using a digital thermometer, and disturbed soil samples were used to determine soil moisture. For moisture quantification, the oven-drying method was used: a representative soil sample was collected, weighed to obtain its wet mass, and then placed in an oven set at 105–110 °C. The samples were dried for 24 h or until the weight stabilized, indicating the complete removal of water. After drying, the samples were cooled in a desiccator to prevent moisture reabsorption and reweighed to obtain the dry mass. Soil moisture content was calculated using the following formula [28]:

$$\mathrm{M}\mathrm{o}\mathrm{i}\mathrm{s}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{e}\left(\mathrm{\%}\right)={\displaystyle \frac{(\mathrm{W}\mathrm{e}\mathrm{t}\mathrm{ }\mathrm{M}\mathrm{a}\mathrm{s}\mathrm{s}-\mathrm{D}\mathrm{r}\mathrm{y}\mathrm{ }\mathrm{M}\mathrm{a}\mathrm{s}\mathrm{s})}{DryMass}}\times 100$$

Microbial biomass-C was obtained using the microwave-assisted extraction method [30], which uses electromagnetic energy (microwaves) to facilitate energy transfer and temperature increases, resulting in cell lysis and the release of intracellular compounds. Based on the method established by [31], the energy required to induce bacterial lysis is 800 J s⁻¹ g⁻¹ of dry soil. The formula used to calculate the irradiation time is given by:

$$\mathrm{I}\mathrm{r}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\mathrm{T}\mathrm{i}\mathrm{m}\mathrm{e}(\mathrm{s})={\displaystyle \frac{\left[Desiredincidentenergy\right(J)\times Massofdrysoil\left(g\right)]}{MicrowavePower\left(W\right)}}$$

Following the irradiation, extraction was performed using 0.5 mol L⁻¹ potassium sulfate, oxidation was conducted with 0.066 mol L⁻¹ potassium dichromate, and titration was conducted with 0.033 mol L⁻¹ ammonium ferrous sulfate. This methodology provides a reliable and efficient approach for assessing the MBC in soil samples.

The option for collar fixation in soil and readings within 30 min is standardized for readings made immediately upon arrival in the field. However, readings taken 30 d after fixation were justified by the effect of soil stabilization prior to recording readings.

2.4. Data Analysis

Six replicates were randomly distributed for each pasture management practice: three replicates each for the MC and three replicates for the FC.

To determine the statistical difference between the FC and MC in evaluating CO₂ flux emissions, monthly collections were considered repetition. Normality and homogeneity of variance were verified using the Lilliefors and Bartlett tests. Data were analyzed using an F-test (ANOVA) (p ≤ 0.10) using the SISVAR Software [32], and means between treatments were compared using the Scott–Knott test (p ≤ 0.10).

The correlation coefficient (r) between CO₂ flux measured in the FC and MC (p ≤ 0.01 and 0.05) was calculated without considering the collection time. Means and standard deviations were calculated for each measurement during each month of data collection.

Mean values of soil moisture, temperature, and microbial biomass were calculated for each treatment and biomass-C. The same variables were calculated as the mean, mean-standard deviations, and coefficient of variation between the collection times for each pasture management practice. Finally, a regression analysis was conducted to evaluate the relationship between CO₂ emissions (μmol m⁻² s⁻¹) and the variables of temperature (°C) and soil moisture (%). Data from different pasture management practices and collection periods over 6 months were used, using PVC collars fixed for 30 d (FC) and 30 min (MC). The statistical significance of these relationships was analyzed using an F-test at a significance level of 5%. Subsequently, the best-fit model was used.

3. Results

3.1. CO₂ Flux Emissions Between FC and MC

Figure 2 and Figure 3 illustrate the CO₂ fluxes associated with various pasture management practices. Using an MC 30 min prior to data collection resulted in significantly higher CO₂ emissions than deploying an FC installed 30 d in advance. Statistical analysis revealed significant differences (p ≤ 0.10) in mean CO₂ emissions between the two collar types. Across all the evaluated pasture management practices, the MC consistently produced greater CO₂ emissions, with notable statistical differences observed in the CON, BUR, and iCL practices (Figure 2 and Figure 3).

Although the FER, PH, and CHI did not show statistical differences (p ≤ 0.10) regarding the CO₂ flux emissions in the MC compared to the FC, the elevation of emissions were 35.39%, 29.25%, and 19.50% in these respective pasture management practices compared to using the FC and the MC (Figure 2 and Figure 3).

Overall, soil management practices involving tillage (both total disturbance, as seen in the PH management approach, and minimal disturbance as in CHI management) resulted in the highest average CO₂ emissions. This pattern persisted regardless of whether the MC or FC was used for measurement. Notably, the area under fertilization exhibited the highest CO₂ emissions, recorded at 3.06 μmol m⁻² s⁻¹ (Figure 2 and Figure 3).

Specifically, fertilization management practice (FER) yielded the highest emissions when measured with an MC fixed 30 min after installation. However, the emissions decreased significantly to 1.26 μmol m⁻² s⁻¹ when the collar was fixed 30 d prior to measurement. Similarly, conservation management practices (iCL) produced emissions of 1.64 μmol m⁻² s⁻¹ (Figure 3).

These findings highlight the influence of the collar installation timing on CO₂ flux measurements, indicating that allowing more time for soil stabilization can lead to lower emission readings.

3.2. CO₂ Flux Emissions During 6 Months of Evaluation

Although CO₂ flux emissions were overestimated when the MC was used compared with the FC, the behavior of the CO₂ flux for each month in the quantification of these gases was similar (Figure 4). The correlation coefficients (r) between the two forms of measurement (fixed and mobile) were 0.91 (p ≤ 0.01) in the PH management and 0.49 (p ≤ 0.05) in the CHI management.

Regardless of the pasture management practice, data collections 3 (May/2014), 4 (June/2014), and 5 (July/2014) were the ones that presented the lowest CO₂ flux emission values. However, collection 1 (March/2014) was the one that presented the highest values (Figure 4).

In the different pasture management practices (CON, CHI, FER, BUR, PH, and iCL), considering both the FC and MC, emissions exhibited a general decline over the evaluation period, with a slight increase observed in the final month of assessment (Figure 4).

3.3. Soil Attributes and Relation with CO₂ Flux Emissions

The highest CO₂ flux emissions in collection 1 (March/2014) were accompanied by higher soil temperature values, regardless of management (

Table 1. Mean values of microbial biomass-C, soil temperature, and soil moisture at 0–5 cm depth after six months of consecutive evaluation under different pasture management practices.

Pasture Managements	Months of Evaluation						$\overline{\mathit{x}}$	Σ	CV
	1	2	3	4	5	6
	Microbial biomass-C (µg g−1 de solo)								(%)
Control	19.15	94.39	116.1	76.15	0.02	73.25	63.18	35.73	56.55
Chisel	166.90	65.66	149.85	102.41	54.4	83.36	103.76	36.41	35.09
Fertilized	138.17	91.66	260.55	139.18	56.29	146.51	138.73	43.35	31.25
Burned	346.10	50.62	168.75	98.48	59.46	27.79	125.20	88.15	70.41
iCL	47.88	15.05	70.20	94.54	21.51	59.36	51.42	23.28	45.26
PH	82.08	47.88	147.15	110.29	86.02	93.46	94.48	22.83	24.16
	Temperature (°C)
Control	27.33	24.67	23.67	24.00	22.67	21.33	23.95	1.39	5.80
Chisel	28.33	25.00	23.33	22.67	22.33	22.67	24.06	1.74	7.23
Fertilized	29.00	26.00	23.67	22.33	23.00	24.33	24.72	1.85	7.49
Burned	31.33	26.33	24.67	23.00	24.33	23.67	25.56	2.18	8.54
iCL	30.00	26.33	23.67	23.33	24.33	26.33	25.67	1.89	7.36
PH	31.00	27.00	23.33	22.33	23.33	23.67	25.11	2.59	10.33
	Soil moisture (%)
Control	7.42	20.05	12.77	11.55	5.90	5.31	10.50	4.29	40.86
Chisel	12.47	24.8	13.86	12.25	5.53	9.07	13.00	4.22	32.49
Fertilized	14.01	25.56	15.21	11.56	7.74	6.69	13.45	4.81	35.75
Burned	10.31	23.05	10.20	9.22	7.64	6.37	11.13	3.97	35.69
iCL	6.58	18.34	11.95	9.08	4.84	6.86	9.61	3.69	38.42
PH	12.37	25.75	12.31	12.83	5.76	9.28	13.05	4.23	32.44

$\overline{x}$—mean values obtained in six months of data collection; σ—mean standard deviation; CV—coefficient of variation. Months of evaluation: 1 March/2014; 2 April/2014; 3 May/2014; 4 June/2014; 5 July/2014; 6 August/2014. iCL: integrated crop-livestock; PH: plowing and harrowing.

).

The highest value of temperature was verified in the burned (BUR) (31.33 °C) pasture management practice, followed by the PH (31.00 °C) management and integrated crop-livestock (ILP) (30.00 °C) management, both observed in Collection 1 (March/2014). The lowest value of soil temperature was verified in the control (CON) (21.33 °C) pasture management practice in Collection 6 (August/2014), followed by the fertilized (FER) and CHI pasture managements, both in Collection 4 (June/2014), which showed a mean temperature value of 22.33 °C. In the PH pasture management practice, the highest values of CV (%) were verified, and in the other pasture management practices, the CV (%) was <8.54%.

For BUR and CHI pasture managements, as with CO₂ flux emissions, the highest values of microbial biomass-C in these areas were observed during Collection 1 (March/2014). The values were 346.10 and 166.90 μg C g⁻¹, respectively, overall being the highest value. The lowest values were observed for the iCL pasture management practice (Table 1).

As shown in Figure 5, irrespective of the management, CO₂ flux emissions showed a significant increase in the temperature and soil moisture (F ≤ 0.10). The 10 °C increase in the soil temperature from 20 to 30 °C, increased the CO₂ flux emissions by 3.52 μmol m⁻². s⁻¹ when the MC was used. In areas that used the FC, the increase was 3.05 μmol m⁻²·s⁻¹. With the increase in soil temperature from 30 to 40 °C, the CO₂ flux emissions increased by 2.84 μm m⁻²·s⁻¹ in the MC and 1.89 μm m⁻²·s⁻¹ in the FC.

With an increase in soil moisture from 5 to 15%, CO₂ flux emissions raised 0.84 μm m⁻²·s⁻¹ in the FC and 1.56 μm m⁻²·s⁻¹ with the use of the MC. When the soil moisture increased from 15% to 30%, the use of the FC had a greater influence, increasing the CO₂ flux emissions to 1.67 μm m⁻²·s⁻¹, whereas with the MC, the elevation was 0.91 μm m⁻²·s⁻¹ (Figure 5). No significant regression coefficients were observed between the biomass-C and CO₂ fluxes.

4. Discussion

The higher values of CO₂ flux emissions measured with the use of the MC compared with the FC are justified by the breakage of aggregates, root rupture at the moment of collar fixation, and intra-aggregate organic matter exposure with high lability [33]. This procedure (MC vs. FC) may have stimulated the release of CO₂ in these compartments through heterotrophic respiration of the soil, thus leading to overestimations in the flux measures.

Physically, under natural soil conditions, almost all the CO₂ produced is transported to the soil surface and released by diffusion, which is a slow process [34,35]. However, when adopting a practice such as collar fixation 30 min (MC) prior to measurements, an alteration of this natural condition was observed, increasing the amount of flux gas exchange by the total pressure gradient, called mass flow, which is a faster process than diffusion, leading to overestimation of data collection.

The advantage of collar fixation 30 d (FC) prior to data collection is the elapsed time, which is typically sufficient to stabilize local soil changes. [36] described that, with a maximum period of 40 d, the reestablishment of the soil structure can be observed under the influence of organic compound. The author of [37] described that because the soil surface layer is more subject to cycles of wetting and drying, soil aggregation may increase in this region. Restoration of the soil structure may be prevented, especially at the collar border, where lateral diffusion ceases to occur, which may be a source of measurement error [38]. In addition, the elapsed time was sufficient for the re-establishment of the ruptured root tissue.

The highest emission values were observed with the use of the FC compared to the MC in the CHI pasture management practice in Collection 2 and FER pasture management practice in Collection 3, possibly because of the higher soil moisture conditions at the time of these collections (Table 1). The elevation in soil moisture promoted an increase in its plasticity, which may have promoted a better adjustment of the soil with the surrounding soil collar, thereby providing better isolation and avoiding lateral flow, as observed in other collections using the MC. In addition, this effect may be related to the application of liming and fertilization or to disturbances in the root system.

The highest values of temperature observed in Collection 1 agreed with the month of higher atmospheric temperature (March) among the months studied, directly influencing all pasture management practices, independent of the collar. The management practices that included some practice of burning or revolving the soil, such as BUR, iCL, and PH pasture managements, were more affected (Figure 2 and Figure 3). Collection 1 also provided the highest values of CO₂ emissions, showing the influence of temperature on carbon emissions in the soil, regardless of whether the collar was fixed or mobile (Figure 4).

According to a previous study [39], for every 10 °C increase, the decomposition activity increases by 2.6-fold for leaves and 2.0-fold for roots; in that way, the CO₂ emission values can be increased. The increase in CO₂ emissions with increasing temperature and soil moisture corroborates studies by other authors [40,41], who indicated that these variables are the most important factors influencing CO₂ emissions in the soil and the decomposition of organic soil matter [42]. In addition, these attributes were identified as attributes with great temporal and spatial variability that were directly related to climatic conditions.

Among the pasture management practices, the highest values of temperature verified in the BUR, iCL, and CHI pasture management practices were justified because of the greater soil disturbance in these management practices. The burning of the pasture in the BUR pasture management practice, the superficial soil layer revolving in the total area in the PH pasture management practice, and the partial revolving soil in the iCL led to changes in the soil microclimate, which directly influenced the litter of the soil by incorporation in the revolving soil or by the combustion in the burning; this promoted local change and lead to the elevation of soil temperature.

The results of the present study demonstrate that fixation of the collar during data collection promotes elevation of CO₂ emissions in comparison to fixation performed prior to it. Although the observed trend is similar, with the possibility of fixing the soil collar for CO₂ measurement 30 d prior to field data collection, the quality of the collected data can be improved by bringing the results of the observed reality into the field. Although the present study contributes to the data collection regarding the emissions of CO₂, collar fixation should be tested to reduce the time between collar fixation and the starting time of data collection in the field.

5. Conclusions

The results of this study showed that fixing the collar 30 min prior to data collection resulted in an overestimation of approximately 32.7% in CO₂ emissions compared to fixing the collar 30 d prior. This significant difference can be attributed to the physical disturbance of the soil that occurs during the installation of the MC, leading to the breakdown of aggregates and exposure to physically protected organic matter, which potentially stimulates microbial respiration in the soil and alters natural conditions for gas diffusion.

Choosing to fix the collar for a longer period (e.g., 30 d) allows the soil to stabilize, promoting a better representation of CO₂ emissions under natural conditions. This is particularly relevant because the elapsed time is sufficient for the damaged root tissue to recover and for the soil structure to re-establish itself, minimizing the measurement errors associated with lateral gas diffusion.

The use of the FC for CO₂ flux measurements should be carefully planned, and future studies should explore different collar fixation periods to enhance the accuracy of the collected data and better reflect the reality of the field.