Carbon and carbon dioxide accumulation by marandu grass under nitrogen fertilization and irrigation ¹

Acúmulo de carbono e dióxido de carbono pelo capim-marandu sob adubação nitrogenada e irrigação

Concern over global climate change has been growing worldwide. Global warming results mainly from the emission of CO₂ and other greenhouse gases (GHGs) such as methane (CH₄) and nitrous oxide (N₂O), leading to the search of strategies to reduce the sources of greenhouse gas emissions, e.g. C sequestration by vegetation (Carvalho et al., 2010; Cerri et al., 2007a; Cerri et al., 2009).

In pastures, C assimilation is directed towards the production of fiber and forage (Luyssaert et al., 2008). Thus, the intensification of the production system, through fertilization and irrigation, can be an important tool to mitigate GHGs in this ecosystem (Conant et al., 2001). According to FAO (Food and Agriculture Organization, 2011), Brazil has 196 million hectares of grasslands, and approximately 80% of this area is in some stage of degradation (Marchão et al., 2007), due to, among other reasons, low soil fertility and lack of nutrient replacement through fertilizer application. For comparison, the annual CO₂ sequestration in pasture conducted extensively vary from 4 to 10 t ha^-1 in the shoot, while in tropical forage, well-nourished pastures, these values ​​range from 30 to 50 t ha^-1 (Primavesi et al., 2007).

Forage grasses adapted to soil and climatic conditions and of high dry matter productivity (DMP) should be used, as they contribute positively in CO₂ sequestration (Conant et al., 2001). In this sense, Brachiaria grasses stand out for being highly productive (Bauer et al, 2011) and highly responsive to fertilization (Rezende et al., 2011; Cabral et al., 2012; Cunha et al., 2012), which is extremely desirable to maximize C sequestration and consequently to mitigate GHGs (Conant et al., 2001).

According to Primavesi et al. (2007), 60% of tropical soils undergo water stress, which can cause severe photosynthesis inhibition (Pinheiro & Chaves, 2011), and 36% of tropical soils have low nutrient reserves, making irrigation and fertilization of pastures in these areas essential. Nitrogen fertilization stands out in this context for being the most limiting nutrient to plant growth (Rezende et al., 2015). Lack of N can directly affect photosynthesis, by affecting the synthesis and activity of the enzyme ribulose 1,5-bisphosphate carboxylase (RUBISCO), responsible for CO₂ assimilation (Pinheiro & Chaves, 2011), and indirectly the GHGs mitigation, as the C sequestration by shoots is compromised by the reduction of photosynthesis (Primavesi et al., 2007).

Therefore, this study aimed to evaluate the accumulation of carbon (C) and carbon dioxide (CO₂), based on dry matter productivity (DMP) and organic matter content (OM) of marandu grass (Brachiaria brizantha Stapf. cv. Marandu) under N rates and irrigation in the Cerrado Paulista during the rainy and dry seasons.

The experiments were conducted in a Typic Dark soil, eutrophic, sandy texture (Empresa Brasileira de Pesquisa Agropecuária - EMBRAPA, 2013), at the UNESP - Universidade Estadual Paulista "Júlio de Mesquita Filho" in Teaching and Research Farm, Ilha Solteira - SP, located on the left bank of the Parana River (20º 21' S and 51º 22' W), 326 m altitude, in an area previously occupied by Panicum maximum Jacq. cv. Colonião undergrazed. Table 1 shows the soil chemical analyzes carried out during the experiment.

The experimental areas were prepared by conventional system with one plowing and two disking, then sowing of marandu grass. At sowing, 20 kg ha^-1 N were applied as urea (45% N); phosphorus (superphosphate - 18% P₂O₅) and potassium (potassium chloride - 60% K₂O) were provided to adjust the phosphorus levels at 30 mg dm^-3; and potassium occupying 5% of the cation exchange capacity (CEC).

Experiments were conducted with and without irrigation (Figure 1), and the N fertilization was evaluated during the rainy and dry seasons. The irrigated experiment used a fixed sprinkler irrigation system with nozzles spaced 12 × 12 m, with an average rainfall of 7.0 mm h^-1 and mean Christiansen's uniformity coefficient of 84.5%. The irrigation had a 3-day schedule with replacement of the reference evapotranspiration estimated by Penman Monteith (Allen et al., 1998) and crop coefficient of 1.0.

Five N rates (0, 50, 100, 150 and 200 kg^-1 ha^-1 after each cutting) were applied as urea, resulting in the application of 0, 300, 600, 900 and 1200 kg ha^-1 in the rainy season and 0, 100, 200, 300 and 400 kg ha^-1 in the dry season. The experiments were arranged in a split plot randomized block design, with plots represented by N rates and the subplots by seasons (rainy and dry seasons). The plot areas were 9.0 m² (3 x 3 m) spaced 2 m apart.

After the initial growth, the first cutting of marandu grass was performed and N fertilization was applied to the subplots. Every after three cuttings 18 kg ha^-1 K₂O (potassium chloride) were applied, and a new soil chemical analysis was performed to adjust the phosphorus again to 30 mg dm^-3, with potassium occupying 5% of CEC, in order to meet phosphorus and potassium plant requirements and optimize the N response. Five months after the start of N fertilization, 81 kg ha^-1 P₂O₅ (superphosphate) and 96 kg ha^-1 K₂O (chloride potassium) were applied to the plots, and seven months later, 71 kg ha^-1 P₂O₅ (superphosphate) and 130 kg ha^-1 of K₂O (potassium chloride) were also applied.

Cuttings were performed manually, 15 cm above the ground in the center of the subplots, in an area of 1.0 m². The intervals between cuttings ranged from 28 to 32 days in the rainy season, and 40-45 days in the dry season. The remaining biomass in the area was mechanically mowed, removed from the plot, and then N was applied to the subplots.

The clipped forage was packed in paper bags and then incubated in a forced circulation oven at 60 °C, for 72 hours, to determine dry matter content. Then, the samples were ground in a Wiley mill, with a 1 mm mesh screen, and analyzed for the content of organic matter (total carbon) following the method described by Silva (2002).

Dry matter productivity (DMP) was calculated by multiplying the amount of green mass (kg m^-2) by the content of the initial dry matter, and then extrapolating to one hectare. The estimates of OMP, C and CO₂ sequestration were based on the DMP and OM content (%) (Table 2), using the respective equations described by Primavesi et al. (2007): OMP (t ha^-1) = DMP x OM%; C (t ha^-1) = OMP / 1.724 and CO₂ (t ha^-1) = C (t ha^-1) x 3.67.

Analyses of data separated and joined were carried out after verification of normality and homogeneity of variances, and by applying the analysis of variance (F test). When significance was found, regression analysis was performed for N rates and mean comparison for the seasons (Tukey test at 5% probability). Next, a combined analysis of the irrigated with the non-irrigated experiments was performed through the comparison of means (Tukey test at 5% probability) of the factor irrigation using the Statistical Analysis System (SAS, 2004).

There was no significant interaction (p > 0.05) between N and seasons in the irrigated experiment due to the inexistent water restriction, which is the main limiting factor for DMP in the dry season. However, N supply provided a quadratic increase in OMP and C and CO₂ sequestration by marandu grass (Figure 2). It is known that N is the nutrient that most influence plant growth (Rezende et al., 2015), increasing plant DMP linearly up the point where other growth factors become limiting. Thus, OMP and sequestration of C and CO₂ reflected the effect of N on DMP when there was adequate water availability.

Sequestration of C above 1 t ha^-1 and CO₂ above 5 t ha^-1, per cutting, by marandu grass with N supply was verified, demonstrating that N fertilization associated with irrigation is essential to increase GHGs sequestration by pastures in a first moment. Pinheiro & Chaves (2011) pointed out that the lack of N can directly affect photosynthesis, by affecting the synthesis and activity of the enzyme responsible for the assimilation of CO₂ (RUBISCO), limiting the accumulation of C and CO₂ by the vegetation. However, it is essential to adopt the correct management of N fertilization because of the emissions of N₂O (and other GHGs) arising from the use of N fertilizers (Barcellos et al., 2008; Virkajarvi et al., 2010) when the source and the form of application recommended for each situation are disregarded.

In the non-irrigated experiment, the interaction between N rates and seasons was significant (p < 0.05) only for the rainy season. This result can be attributed to a higher rainfall recorded in the rainy season (> 900 mm) compared with the rainfall recorded in the dry season (< 200 mm) (Figure 1). The primary ion-root contact mechanism in the absorption of N is the mass flow, which is dependent on water (Malavolta et al., 1989). Therefore, low water availability may limit the plant metabolism, uptake of N and, consequently, the plant productive response in terms of production (Primavesi et al., 2007, Pinheiro & Chaves, 2011). According to Lopes et al. (2005), the DMP and, consequently, the OMP, the sequestration of C and CO₂ by grasses during the dry season can be 70% lower than in the rainy season. The highest OMP (3686 kg ha^-1) and sequestration of C (2136 kg ha^-1) and CO₂ (7846 kg ha^-1) were recorded with application of 162 kg ha^-1 N in the rainy season (Figure 3).

Carbon dioxide sequestration per cutting (7846 kg ha^-1) was extremely high and confirms the beneficial effects caused by N fertilization and irrigation compared with the range between 30 and 50 t ha^-1 annual cited as indicative of well-managed pastures by Primavesi et al. (2007). Assuming that the climatic conditions remained constant, the final CO₂ accumulation in the rainy season would be approximately 47 t ha^-1. Paulino & Teixeira (2009) reported that the lack of N fertilization and less frequent grazing resulted in loss of 57 g of C per square meter per year to the atmosphere. This report confirms the importance of N fertilization on GHGs mitigation.

Table 3 shows the effect of season in OMP, accumulation of C and accumulation of CO₂ by marandu grass in the non-irrigated experiment. There was higher (p < 0.05) OMP and sequestration of C and CO₂ in the rainy season due to increased water availability (> 900 mm) compared with the dry season (< 200 mm) (Figure 1). The water stress affects various physiological processes of plants, generally increasing the stomatal conductance, reducing perspiration and hence the CO₂ supply to perform photosynthesis (Taiz & Zeiger, 2009). Therefore, biomass production and C sequestration by plants are compromised in environments with low water availability, even when there is adequate nutrient supply.

Loss et al. (2013) evaluated the effect of irrigation in the stock of total organic carbon (TOC) of an Oxisol with Cynodon grass and found that the irrigated area had the highest TOC values, regardless of the depth of evaluation. The authors attributed the result to higher production of green mass, under irrigation. It is noteworthy that the stock of C in the soil is derived in part from the decomposition of plant residues, which at some point captured CO₂ from the atmosphere to produce biomass (Ministry of Agriculture, Livestock and Supply - MAPA, 2011; Matheus, 2012). Thus, the results obtained by Loss et al. (2013) confirm the benefit provided by irrigation in the mitigation of greenhouse gases.

At the end of the trial period, not considering the effect of season and N rates, the factor irrigation was predominant to maximize the OMP and sequestration of C and CO₂ by marandu grass (Table 4). Irrigation provided increments of approximately 20% in sequestration of C and CO₂ compared with the non-irrigated experiment, contributing considerably to the mitigation of GHGs.

The Brazilian herd contingent associated with the large area occupied by pastures point out the importance and impacts of livestock activity such as GHGs emitting source or as a drain, depending on the management used (Saussana et al., 2007). It is therefore essential to optimize the conditions for the cultivation of grasses, since under suitable vegetable mass production, grasses are able to sequester substantial amounts of C fixing it to the soil in the organic form (Cerri et al., 2007b).

Nitrogen supply and irrigation increases the potential sequestration of C and CO₂ by marandu grass, and this increase is higher during the rainy season.