The Atlantic Forest is an important Brazilian biome, as it is one of the forests with the greatest biodiversity on the planet and presents a high level of eudemism (Souza et al., 2021). It stands out for its coverage of 24.3% (32.9 million ha) of primary and secondary forests (MapBiomas, 2022), with forest fragments of up to 50 ha (Scarano & Ceotto, 2015), surrounded by anthropogenic matrices, pastures and agricultural areas (Joly et al., 2014).

Among these areas, forests located in urban areas (called urban forests) can be conceptualized as networks or systems that comprise all forests, groups of trees and individual trees located in urban and peri-urban areas (Brun et al., 2017). They may be formed by green areas containing remnants of natural landscapes with different land use histories (Elmqvist et al., 2013). According to Mapbiomas data (2024), 6.9% of urban areas in Brazil are covered by vegetation, amounting to 283,700 ha. Of this total, 61.5% (174,599 ha), are in the Atlantic Forest biome. In other Brazilian biomes, approximately one in every five ha (22%) is in the Cerrado, with 62,533 ha of vegetation. The remaining 16.5% is distributed among the Amazon (18,605 ha – 6.6% of the total), Caatinga (16,139 ha – 5.7%), Pampa (11,228 ha – 4%), and Pantanal (587 ha – 0.2%). These areas in the Atlantic Forest are essential, as they are related to the quality of life and health of approximately 120 million people, or 72% of Brazil’s population, according to SOS Mata Atlântica (2020).

However, such areas are vulnerable to human actions such as deforestation and forest degradation, phenomena responsible for greenhouse gas emissions of approximately 5-10 Gt.CO_2eq (Bhatti et al., 2023), making this biome one of the hotspots with the biggest threats to biodiversity (Romanelli et al., 2022). Despite this vulnerability, urban forests provide ecosystem services to the surrounding population, regulating the microclimate, protecting biodiversity, and improving quality of life (Alonzo et al., 2016). They also contribute to the removal of atmospheric carbon dioxide (CO₂), whose potential has already been recognized and highlighted in scientific research as the largest and most cost-effective carbon sink (Wang et al., 2021). In this context, it is essential to quantify the carbon stored in forest biomass, since biomass and carbon stocks are conditions and controls of the global carbon cycle, serving as indicators for constructing scenarios relating to climate change, as well as defining strategies for mitigating its impacts (Lima et al., 2021). To this end, determining carbon stocks in forests can be conducted in two ways: direct methods (forest inventory and destructive sampling of individuals) and/or estimated indirectly (using allometric equations and data derived from remote sensing technologies) (Romero et al., 2020). In cases where direct determination is possible, some samples (such as 31 trees used by Veres et al., 2019) are often taken to quantify carbon in such a way that these samples generate a database for fitting regression models, which enable estimating biomass and carbon.

Regarding carbon stock estimation, Azevedo et al. (2018) conducted a study to quantify aboveground biomass and carbon stock in reforested areas aged 3, 5, and 7 years in Cachoeiras do Macacu, Rio de Janeiro, Brazil, using allometric equations (indirect methods). The results showed that both biomass and carbon increased with forest age. At 3 years, the biomass was 39.88 t.ha^-1 and the carbon was 19.94 t.ha^-1. At 5 years, these values increased to 45.78 t.ha^-1 and 22.89 t.ha^-1, respectively. At 7 years, the biomass reached 71.24 t.ha^-1 and the carbon was 35.62 t.ha^-1. Similar results were obtained in an area located in Botucatu, São Paulo, Brazil, with biomass ranging from 113.28 t.ha^-1 to 130.87 t.ha^-1 (Pontes et al., 2019).

Thus, there is a need and urgency to develop environmental mechanisms to mitigate the effects of climate change in the Brazilian context, as the country is part of global agreements to reduce greenhouse gases (GHG) emissions such as the Paris Agreement (Koh et al., 2021) and has established a National Policy related to Climate Change (Brazil, 2009), main national legislation on this topic. Therefore, accurate and reliable estimates of forest biomass and carbon must be generated, especially in a biome of such relevance as the Atlantic Forest. In view of the above, the objective of this work was to evaluate the carbon stock and carbon equivalent dynamics in the biomass of forest species located in a native Atlantic Forest remnant in southern Brazil.

According to the results presented, there is vast potential for carbon fixation (64.23 t.ha^-1) in the biomass of forest species existing in the native fragment located in the MOF area, since only the biomass contained in the aboveground compartment (and consequently the carbon) was considered, excluding biomass from other compartments such as necromass or soil. This fact is related to species diversity, a hypothesis supported by the identification of 144 species and 92 genera during the 15 years collection (2006 to 2021). This factor significantly contributed to the increase in carbon stock due to the optimized photosynthesis of the different species present at the site (Catovsky et al., 2002).

Furthermore, the area presents a dominance of early and late secondary species belonging to the following genera: Allophylus, Araucaria, Casearia, Citronella, Clethra, Gymmanthes, Jacaranda, Luehea, Moquiniastrum, Myrcia, Ocotea, Schinus and Symplocos. The species succession stage ha a direct interaction with carbon storage, as the initial and medium stages, which are in the growth phase, accumulate greater amounts of biomass and carbon (approximately 42.11 t.ha^-1 of biomass) (Veres et al., 2019).

Over the 15-year monitoring period, there was a 30% increase in the biomass of the analyzed fragment, suggesting that species richness played an important role, particularly due to the different growth rates of each species and the consequent accumulation of biomass. Species diversity was also an important factor in the research by Pontes et al. (2019) and Capellesso et al. (2021). These authors highlighted that the structural characteristics of trees, the density of the forest, the forest's resilience and stability, as well as different successional stages and life cycles, impacted the composition of the plant community and the capacity for carbon absorption.

This pattern was also evident in an area located in MOF in General Carneiro, Paraná, Brazil, with an increase of 0.75 t.ha^-1.year^-1 for biomass and 0.31 t.ha^-1.year^-1 for biomass carbon (Mognon et al., 2013). The increase of approximately 30% in biomass also occurred in an area with different successional regeneration stages in the Northwest Region of the State of Rio Grande do Sul, Brazil (Erthal et al., 2023), reaffirming the carbon fixation potential of native forest fragments of the South of the country (Machado et al., 2019).

Growth and entry rates of individuals were higher than mortality rates during the monitored period, except for 2015-2018. The entries directly contributed to floristic enrichment and species diversity in the area, and growth was influenced by the satisfactory climatic and environmental conditions of the site, as evidenced by Batista et al. (2020). Furthermore, extreme climatic and meteorological conditions, such as prolonged droughts, storms, heatwaves, and frosts, can increase mortality and reduce tree growth, as can changes in competition for resources (water, nutrients, light, among others).

Entries over time significantly contributed to biomass accumulation, as according to Caron et al. (2015), the greater the density of individuals in an area, especially with advanced successional stages, the greater the biomass production per unit area.

The carbon removal potential in the study area during the 15 years of monitoring (2006 to 2021) was 64.23 t.ha^-1, with an average of 4.06 t.ha^-1.year^-1, being similar to other studies conducted in similar regions, such as that by Souza et al. (2023), conducted in a hotspot of the Atlantic Forest, in the Mantiqueira Range, Southeastern region of Brazil. In this study, the authors assessed the carbon stock and uptake patterns of the sampled forests along a high-altitude gradient (1500–2100 m a.s.l.) and monitored in two inventories (2011 and 2016). The results indicated variations in carbon accumulation over the period, with a carbon gain of 3.82–5.14 t.ha⁻¹.year⁻¹.

The study conducted by Reis et al. (2019) quantified the carbon stock of 600 trees of five species (Tipuana tipu (Benth.) Kuntze, Acer negundo L., Ficus benjamina L., Terminalia catappa Linn, and Licania tomentosa (Benth.) Fritsch planted in three Brazilian cities (Curitiba, Paraná; Itanhaém, São Paulo; and Bonito, Mato Grosso do Sul, Brazil) using the allometric equations of Brown (1997) and Brianezi et al. (2013). According to the Brown (1997) equation, the species with the highest occurrence in Curitiba, Paraná (A. negundo. and T. tipu) had, respectively, 20.61 tons of carbon and 75.64 tCO_₂eq, and 95.17 tons of carbon and 349.27 tCO_₂eq. Using the Brianezi et al. (2013) equation, 21.58 tons of carbon and 79.21 tCO_₂eq were found for A. negundo, and 88.51 tons of carbon and 324.83 tCO_₂eq for T.a tipu.

According to the obtained results, individuals of A. angustifolia with a diameter greater than 60 cm exhibited higher carbon allocation. This is due to the characteristics of the analyzed fragment, which shows a higher occurrence of individuals of this species compared to trees of other species with smaller diameters and in earlier stages.

This behavior was also observed by Lipinski et al. (2017) in São João do Triunfo, State of Paraná, Brazil, in the analysis of the temporal and spatial dynamics of biomass and carbon between 1995 and 2012. Araucaria angustifolia was predominant throughout the period, with biomass stocks of 102.5 t.ha^-1 in 1995; 126.8 t.ha^-1 in 2012; 43 t.ha^-1 of carbon in 1995; and 54 t.ha^-1 in 2012, revealing an increase in 1.17 t.ha^-1.year^-1.

It is important to highlight those factors such as tree density, size of individuals, species composition, and canopy architecture (denser and symmetrically shaped canopies), as pointed out by Brianezi et al. (2013) and Nowak et al. (2013), can influence the carbon absorption capacity of forests.

In addition to the factors mentioned, it is important to highlight the differences between urban forests and natural forests, which can influence the dominance of A. angustifolia. In urban environments, forests are susceptible to anthropogenic actions and land management practices, which can alter forest structure and species diversity, favouring the establishment of species with greater tolerance to urban stresses, while native or dominant species in natural forests may be less prevalent or replaced by other species more adapted to the urban environment (Olgun et al., 2024). The species A. angustifolia, native to the Atlantic Forest in southern Brazil, may not exhibit the same dominance in urban forests as it does in natural forests because, in urban forests, Araucaria is subject to competition with exotic species and may have a heterogeneous distribution (Olgun et al., 2024). Moreover, A. angustifolia is a large, long-lived species that contributes significantly to carbon storage, and over time, due to natural forest dynamics, it may be replaced by other species with different growth characteristics and life cycles, potentially altering the carbon storage potential of the forest. Thus, the diameter at breast height distribution may suggest the presence of many small trees and few large trees, reflecting a successional stage where carbon stocks might be lower due to the reduced biomass of large trees (Chazdon, 2008). Furthermore, there is a gap related to the development and adjustment of allometric equations for urban forests to estimate the variables analyzed in this study. This occurs due to the impossibility of directly determining biomass, which leads to the use of adjusted equations developed for established native forests, which can cause overestimations in trees located on urban roads.

The evaluated remnant located in an area of MOF in the Atlantic Forest biome revealed that an average of 4.06 t.ha^-1.year^-1 of carbon were removed, with Curitiba emitting 1.85 tCO_2eq per inhabitant, according to the Inventory of Greenhouse Gas Emissions in the city of Curitiba, base year 2016 (Curitiba, 2019). Therefore, it is noted that the evaluated fragment positively contributes to combating climate change, being capable of removing twice the amount of carbon emitted.

Finally, it is recommended to investigate the role of urban forests in the face of climate change, especially in reducing GHG emissions and obtaining information about the carbon stock contained in the biomass of these areas. Furthermore, it is essential to assess the conservation status of forest fragments located in urban areas to conserve existing biodiversity, and if necessary, to restore the site (Batista et al., 2020).

Based on the assessment of the potential for removing CO₂ from the atmosphere by a native urban forest fragment located in the Atlantic Forest biome between 2006 and 2021, the following conclusions can be drawn regarding i) overall dynamics, i.e., the entire assessment period (2006-2021); ii) dynamics by genus (considering that there are different species in different successional stages) and by diameter classes (to illustrate differences in carbon accumulation).

General: the 92 genera and 144 species identified had a positive biomass, carbon and equivalent carbon dioxide balance during the evaluation period. The fragment removed an average of 4.06 t.ha^-1.year^-1, and produced 156.56 t.ha^-1 of biomass, accumulating 64.23 t.ha^-1 of carbon and promoting the removal of 235.51 t.ha^-1 of CO_2eq from the atmosphere, acting to reduce greenhouse gas emissions.

Genus: The 13 predominant genera during the evaluation period were responsible for storing 301.20 t.ha⁻¹ of carbon, or 78.16%, highlighting the largest contribution to carbon sequestration by the following genera: Araucaria, Ocotea, Luehea, and Casearia, with 104.15 t.ha⁻¹ (27.03%), 41.20 t.ha⁻¹ (10.69%), 33.85 t.ha⁻¹ (8.78%), and 29.98 t.ha⁻¹ (7.52%), respectively.

Diametric class: Araucaria individuals with diameters ranging between 60 and 70 cm showed higher carbon fixation (34.62 t.ha^-1, equivalent to 33.28%). For other genera, the diameter classes greater than 20 cm and from 20 to 30 cm were responsible for the highest carbon storage (173.21 t.ha^-1), representing 61.04%.