Assuming that the increase in the CO₂ concentration in the atmosphere is the main cause of accelerating climate change, one of the first steps needed to mitigate climate change is to reduce the CO₂ concentration (Yamasaki, 2003). Planting trees is often advocated by governments and non-governmental organizations as one of the simplest approaches for mitigating the increasing CO₂ concentration, as well as for generating strong public support (Kirschbaum et al., 2024). Thus, Brazil is committed to restoring approximately 21 million degraded hectares in permanent preservation areas (PPAs) and legal reserves on rural properties (Soares-Filho et al., 2014; Brancalion et al., 2016).

Nevertheless, in order to achieve the proposed bold goals, it is necessary for the projects to be economically attractive since these kinds of projects have high costs (Brancalion et al., 2019). The Brazilian Forest Code under law 12,651/2012 (Brazil, 2012) requires that 20% of the rural property area in the Atlantic Forest in Brazil be covered by native vegetation as a Legal Reserve (LR). However, Brazilian legislation allows rural landowners to exploit their Legal Reserve areas for economic gain provided that a sustainable management plan is prepared. A maximum of 50% of the planting can be performed with exotic species, and such plans must be approved by the governing body. Based on these requirements, different systems can be used to reduce implementation costs, such as the system presented by Amazonas et al. (2018, 2021). In these systems, Eucalyptus spp. are planted with a density of up to 50% coupled with a high diversity of native tree species, representing a compromise between production and restoration (Amazonas et al., 2018).

The use of Eucalyptus spp. is strategic because it reduces implantation costs, as the cost of Eucalyptus spp. seedlings is low and these plants have high plasticity, resistance to drought, high silviculture technology, and a well-established economic market (Silva, 2017). The use of Eucalyptus spp. trees may also favour understorey regeneration, as observed in other studies (Brockerhoff et al., 2013; Pryde et al., 2015; Wu et al., 2015). However, the use of Eucalyptus spp. may also have some noteworthy disadvantages, such as the possibility of competition with native species and the potential to deplete soil nutrients in the long term. Therefore, it is important to carefully evaluate each project and understand the growth dynamics of species belonging to different ecological groups intercropped with Eucalyptus spp. and their influence on the recovery of degraded areas.

Thus, the objective of this study was to evaluate the growth of native and exotic species, particularly the Eucalyptus genus, for restoring an Atlantic Forest region in Brazil, as well as the carbon stock potential in the aboveground (trunk/canopy) and belowground (root) biomass of the trees. Although Amazon et al. (2018) found a significant effect on the introduction of the Eucalyptus genus in restoration studies, our focus is to quantify its influence on the growth of native species at different planting densities up to 50% of the planting density, as permitted by Law 12,651/2012 (Brazil, 2012). We hypothesise that: (i) the growth and survival of native species will be minimally affected by incorporating different Eucalyptus urophylla x Eucalyptus grandis hybrid densities in the area; (ii) E. urophylla x E. grandis hybrid trees will perform better in treatments with a greater proportion of native species. Despite being a fast-growing species and a strong competitor, Eucalyptus hybrids may exhibit enhanced growth in such conditions due to reduced intraspecific competition.

The experiment was implemented at Fazenda da Lagoa, which is part of the Federal University of Lavras (UFLA) in the municipality of Santo Antônio do Amparo, Minas Gerais (20° 56’40”S, 44°55′8″W), at an elevation of approximately 1,000 m. According to the Koppen climate classification system, the region has a Cwa climate type, which is characterised by tropical and subtropical rainy mesothermic conditions with dry winters and rainy summers (Alvares et al., 2013). The average annual temperature is 19.8°C, and the average annual rainfall is between 1400 and 1700 mm. The soil was characterised as a Dystrophic Oxisol (LVd in the Brazilian soil classification system). The region is composed of a mosaic of Brazilian savanna (Cerrado) and Atlantic Forest phytophysiognomies and is in an agricultural zone. Previous land uses consisted of coffee plantations and pasture.

An E. urophylla x E. grandis hybrid and 18 native species were used for the present study (Table 1). The seeds of the native species were collected in the municipalities of Lavras, MG, and Mariana, MG, and the seedlings were germinated in the Forestry Nursery of the Department of Forest Sciences (UFLA) in the municipality of Lavras, MG, from June 2019 to February 2020. The climatic conditions during germination included an average temperature of 28.3°C and average monthly rainfall ranging from 8.6 to 190.2 mm. Tubes measuring 120 cm³ were used for germination. The E. urophylla x E. grandis hybrid seedlings were produced in 55 cm³ tubes and purchased from a commercial nursery located in the city of Lavras.

After sowing and germination, the native plants were allowed to germinate for 60 days in a shade house (with 50% irradiance and irrigation supplied by microaspersion three times a day for five minutes each at a flow rate of 140 L h^-1). The plants were subsequently transferred to full sun for hardening (irrigation four times a day for five minutes each, with a flow rate of 95 L h^-1) until being transferred to the field.

Leaf-cutting ants and weeds were periodically manually controlled in the planting area. Soil correction was performed by applying 2 Mg ha^-1 of calcitic limestone to supply Ca and Mg. The soil was prepared by harrowing the entire area. Planting was performed in March 2020, with 150 g of formulated NPK 06-30-06 fertilizer per hole. Next, topdressing fertilization was performed in two doses, with 100 g of formulated NPK 20-00-20 being applied each time. Boric acid was applied at a concentration of 10 g per plant in May 2021, which is the beginning of the dry season in the region.

The experiment consisted of a completely randomised block design with four replicates. A 3 × 19 factorial scheme was used, in which the first factor corresponded to the percentage of Eucalyptus plants relative to native species and the second factor was the species that composed the plantation. Two control treatments were also implemented for comparison: one with 100% native species and one with 100% Eucalyptus (Table 1). The plots consisted of nine rows with nine plants each, totalling 81 trees per plot at a spacing of 3 × 2 m (1,667 seedlings per hectare). Additionally, one row of native species was planted between the plot rows to serve as a border area. The arrangement of species in each treatment can be found in Fig. S1, in the supplementary materials.

The planting models were characterised by the proportions of native (pioneer and non-pioneer) and exotic (Eucalyptus) species. Control 1 (T1) had 100% native species; control 2 (T2) had 100% Eucalyptus; model 1 (M1) had 25% Eucalyptus per ha; model 2 (M2) had 33% Eucalyptus per ha; and model 3 (M3) had 50% Eucalyptus per ha.

The percentage of dead plants per species, height (H) and diameter at breast height (DBH) at 42 months after planting were evaluated. The diameters for trees with up to three branches at a height of five centimetres from the soil were converted into fused diameters, as described by Scolforo and Thiersch (2004). Then, the stem cross-sectional area (SA) of each tree was determined from the diameter based on equation (i).

In which: SA is the cross-sectional area (cm² tree¹) and DBH is the diameter at breast height (cm).

In turn, the model described by Binkley et al. (2020) (equation (ii)) was used to obtain the trunk biomass of the Eucalyptus trees. This equation was developed for the clone investigated in this study, E. urophylla x E. grandis, across climatic gradients in Brazil.

In which: y is the biomass in kg tree^-1, DBH is the diameter at breast height (cm), and H is the height (m).

The stem biomass of the native species was calculated from equation (iii), which was obtained by (Luz, 2024) for Atlantic Forest species across climatic gradients in Brazil

In which: B is the biomass (kg tree^-1), DBH is the diameter at breast height (cm), H is the height (m), and ρ is the density (g cm^-³).

The total biomass (B) was obtained using the biomass expansion factor BEF and Root-to-Shoot Ratio (R) values provided by the Intergovernmental Panel on Climate Change (IPCC). The BEF and R values for Eucalyptus were 1.20 and 0.35, respectively, while the BEF and R values for native species were 3.16 and 0.37, respectively. Total carbon was calculated by multiplying the total biomass by 0.47 (IPCC, 2006).

The natural regeneration of the plantation was also evaluated at 42 months by counting each species and determining their occurrence frequencies. Three subplots were established in the upper left corner, centre and lower right corner of each plot. A quadrant of 1 m² was used, and all individual trees greater than 0.5 m in height were counted.

Next, we assessed the normality of the data using the Shapiro-Wilk test at a 5% significance level to evaluate the overall and individual growth potentials of the species in the planting models. An analysis of variance (ANOVA) was conducted following the normality test. When significant differences were detected, the means of the planting models were compared using the Tukey’s test at a 5% significance level, while the means of the species were compared using the Scott-Knott test (5%). Furthermore, the native species and Eucalyptus species groups were analysed separately using ANOVA to better understand the growth behaviours of the different measured variables. The means were again compared using the Tukey’s test when significant differences were identified (5%).

The use of Eucalyptus to regenerate the legal reserve area significantly interfered with the growth of S. granulosoleprosum and E. contortisiliquum, while the models which favoured the growth of native species were M1 and M2 (Fig. 3). This contradicts the first hypothesis because planting Eucalyptus may interfere with the growth of some native species, especially fast-growing species such as S. granulosoleprosum and E. contortisiliquum. It was also observed that a Eucalyptus density of 50% increased the mortality of native species. Similar results were reported by Amazonas et al. (2018), in which Eucalyptus negatively affected the growth of pioneer species (i.e. those with faster growth). Because it is a species which has a rapid initial onset, Eucalyptus aggressively competes for available resources such as water, light and nutrients.

When the cutting cycle of a Eucalyptus tree is complete, the supply of available resources will increase and favour the growth of other native species (Brancalion et al., 2019). Since species with secondary growth grow slower than pioneer species, they require shade to germinate and become established in an area, so the rapid growth of Eucalyptus may aid in developing these species. Thus, Eucalyptus planting densities of up to 33% are recommended for regenerating LR areas so as not to compromise plant quality and to prevent seedlings of native species from being suppressed by Eucalyptus.

Among the different models, the native plants only showed significant differences for DBH, SA, B and C, while the M1 and M2 models showed similar growth patterns. These results once again highlight the potential of using Eucalyptus at the age evaluated with a planting density of 33% for restoring legal reserve areas.

The species which generally showed the greatest growth were classified as pioneers. Increases in the SA, B, and C of Eucalyptus were greater in the M1 and M2 models, which led us to accept the second hypothesis. Eucalyptus is an aggressive competitor and can benefit from the conditions in mixed plantations; it can particularly take advantage of light resources because their canopy is taller than those of other species. Eucalyptus stands out in B and C production, which supports its use in the commercialisation of by products such as sawn wood (at more advanced ages of approximately 12-20 years, as there will be no thinning) and energy, ensuring a reduction in the cost of implementing a restoration project.

Other pioneer species with exceptional performances are S. granulosoleprosum and E. contortisiliquum, which presented greater growth and B stocks than any of the other native species. In addition to the financial gain from commercialising Eucalyptus wood, native species can be exploited in the long term for commercialising economically valuable timber, provided that the limitations of Law N^o. 12.651 are respected (Brazil, 2012), and this approach may make projects eligible for the carbon credit market.

Furthermore, the rapid growth of Eucalyptus is a very important feature for recovering degraded areas and may be an interesting strategy for creating rapid soil cover. In addition, leaf fall increases soil organic matter, favouring nutrient cycling (Carvalho et al., 2017). Oliveira et al. (2014) implemented a mix of native and exotic species (E. grandis and C. citriodora) in different municipalities in the state of Paraná, Brazil, and obtained satisfactory wood production values which could finance the associated forest restoration project. This result confirms that Eucalyptus has great potential to considerably increase and improve the incomes of rural landowners or dampen investments in restoration projects.

This study showed that it is feasible to establish highly diverse mixed plantations with Eucalyptus and native tree species. A Eucalyptus planting density of up to 33% is suitable for restoring legal reserves, as the results showed that this density did not negatively affect the growth of native species up to the evaluated age. Moreover, Eucalyptus seedlings are less expensive than those of native species, which further reduces the costs of implementing restoration projects.

The species with the highest incidence regarding natural regeneration was V. polysphaera, which is considered an aggressive fast-growing species (Dutra et al., 2003) and was especially abundant in T1. This treatment resulted in smaller individuals (Fig. 1) and less soil cover, allowing greater light entry, which benefitted propagation of species that thrive in early successional stages.

Schinus terebinthifolius, P. guajava and A. polycephala were observed in the other planting models. In addition, the reduced number of species found in the area may be related to the fact that weed control was performed up to approximately two years after planting. Despite weed control, Brancalion et al. (2019) observed that areas with seed rain and adequate conditions for germination in restored plantations subjected to intensive silviculture with plants up to two years of age tended to have greater soil cover, leaf area index values, and diversity of regenerants than plantations without intensive silviculture.

Viani et al. (2010) indicated that the richness, density and structure of natural regeneration are influenced by several factors, such as the characteristics of an area and planting age. Thus, it remains necessary to track the development of planted areas over time to understand the quality of the natural regeneration in the area. According to Brancalion et al. (2020), the introduction of Eucalyptus to mixed planting areas with native species favoured natural regeneration throughout the understorey after five years and considerably reduced the costs associated with planting in the restoration area. The authors also observed that Eucalyptus did not regenerate through seed germination in the area, meaning it did not become an invasive species, allowing the area to be recomposed through germination and growth of native species.

The present study only considered growth variables and an initial evaluation of natural regeneration in a transition area between the Cerrado and Atlantic Forest biomes. The species used in this study are also native to the Cerrado, making the results particularly relevant to restoration efforts in both biomes. For future studies, it is important to evaluate additional indicators of restoration success, such as the presence of seed dispersers, seed rain, and the diversity and density of regenerating native species. Furthermore, this project can be adapted to Cerrado-specific restoration efforts, given its relevance to both ecosystems. Additionally, long-term monitoring of plant growth is necessary, as this evaluation was conducted 42 months after planting.

This study demonstrates that incorporating up to 33% Eucalyptus in mixed-species plantations does not significantly affect the growth or survival of native species within the first 42 months. Specifically, proportions of 25% and 33% Eucalyptus increased the diameter at breast height (DBH) and stem cross-sectional area (SA), respectively, while no significant effects were observed on plant height, mortality, biomass accumulation, or carbon storage. These findings support our hypothesis that native species growth would be minimally impacted by the inclusion of Eucalyptus, although its influence should continue to be monitored over time.

The results also confirm that Eucalyptus benefits from reduced intraspecific competition in mixed plantings, aligning with our second hypothesis that it would perform better in treatments with a higher proportion of native species. However, longer term studies are needed to fully understand the dynamics of Eucalyptus-native species interactions, particularly in terms of biodiversity, and the economic potential for carbon credits

In conclusion, our findings suggest that integrating Eucalyptus into reforestation projects can contribute to both ecological restoration and economic objectives, providing a scalable model for Brazil’s reforestation goals, when managed appropriately.