Phenotypic variability and phenological behavior of Caryocar brasiliense under tropical savanna climate conditions

ABSTRACT

This study investigated the phenological dynamics of Caryocar brasiliense cultivated in the northwestern region of Rio de Janeiro state, Brazil, aiming to understand its adaptation to tropical Cerrado conditions and to support its conservation and sustainable management. Biometric variables (height, stem diameter, and crown diameter) and morphological characteristics were monitored and analyzed, in addition to thermal accumulation expressed in degree-days, correlation analyses with climatic factors, and non-linear regression models. The results showed continuous increases in stem and crown diameter, with values comparable to or higher than those of natural individuals, even in young plants, indicating strong growth performance in cultivation. Phenological observations revealed flowering between August and September and fruit abscission concentrated in December, consistent with the local climatic regime. Sprouting showed positive correlations with temperature and precipitation, while defoliation was not influenced by climatic factors. The orchard demonstrated potential for inclusion in intensive fruiting systems, such as high-density planting, an emerging trend in the fruit growing sector. The results highlight the potential for cultivating the pequi tree (C. brasiliense Camb.) in this region, reinforcing its suitability for sustainable production systems and for the recovery of degraded lands in tropical areas. In this context, C. brasiliense shows potential for domestication and use in sustainable production systems, while also highlighting its relevance under climate change scenarios due to its ability to adapt to high temperatures and water-limited conditions.

Key words
pequi; phenology; biometrics; ecological restoration; sustainable fruit production

INTRODUCTION

Tropical savannas, such as the Brazilian Cerrado, are among the most biodiverse ecosystems and play a key role in maintaining ecosystem services and supporting biodiversity. However, the expansion of agriculture and livestock production has led to extensive environmental degradation and vegetation loss, compromising the ecological resilience of these systems (Cunha et al. 2008). In this context, identifying native tree species capable of thriving under variable environmental conditions is essential for both biodiversity conservation and the restoration of degraded tropical landscapes (Fremout et al. 2022).

The pequi tree (Caryocar brasiliense Camb.), widely known in Brazil as the “gold of the Cerrado,” is an emblematic savanna species characterized by perennial growth and drought tolerance (Figueiredo et al. 1989). Its fruit, highly valued for its sensory and nutritional properties, supports local economies through food production and non-timber forest products (Afonso 2022). The mesocarp is rich in lipids, carotenoids, and vitamins, while the species itself provides key ecological functions, including soil protection and food supply for fauna (Santos et al. 2010).

Beyond its ecological and economic relevance, C. brasiliense shows strong potential for integration into sustainable land-use systems, such as agroforestry and restoration-oriented production (Fremout et al. 2022). Because its fruits naturally fall to the ground, harvesting requires minimal inputs, aligning with low-impact management strategies. The phenological cycle and fruit development of perennial species are highly conditioned by environmental factors, such as altitude and temperature, which influence the duration of developmental phases (Colodetti et al. 2025). Despite its broad distribution across the Cerrado biome, few studies have examined the phenological and biometric patterns of the species in ecotonal or transitional regions. Understanding its phenology and growth responses under such conditions is essential to assess its adaptive capacity, identify promising genotypes, and guide sustainable management and restoration strategies, particularly in the context of climate variability.

In recent years, the northwestern region of the state of Rio de Janeiro has experienced increasing climatic instability, characterized by irregular rainfall patterns and elevated temperatures, which intensify water deficits and constrain vegetation establishment and regeneration (Alberton et al. 2023, Macphie and Phillimore 2024). Historically, hilltop areas in this region have been extensively deforested due to agricultural expansion, particularly dairy farming, resulting in degraded environments with low soil fertility and high exposure to adverse climatic conditions (Cunha et al. 2008). These constraints underscore the need for resilient native species capable of tolerating water scarcity and poor soil conditions, while contributing to ecological restoration and sustainable land use (Fremout et al. 2022, Mota et al. 2025). Although C. brasiliense is naturally adapted to the hot and seasonally dry environments of the Brazilian Cerrado, its performance under these specific environmental conditions remains poorly understood, particularly in ecotonal regions beyond its core distribution (Leite et al. 2006, Fagundes et al. 2007). Therefore, it was hypothesized that C. brasiliense can sustain satisfactory growth and phenological performance under such conditions, indicating its potential for use in the restoration of degraded areas and in sustainable production systems in regions subject to climatic constraints.

Accordingly, this study aimed to test the hypothesis that C. brasiliense can maintain satisfactory phenotypic variability, phenological performance, and growth under tropical savanna climate conditions, thereby demonstrating its potential for ecological restoration and sustainable production systems.

METHOD

Description of the experimental area

The experiment was conducted at the Instituto Federal Fluminense, Bom Jesus do Itabapoana Campus, located in the northwestern region of Rio de Janeiro state, Brazil (Fig. 1), at coordinates 21°08’09.8”S and 41°39’34.4”W. According to Köppen’s classification, the local climate is Aw, characterized as tropical with wet summers and a well-defined dry season in winter. The mean temperature in the coldest month rarely falls below 18°C. The orchard was established in 2017 on a hilltop at 88 m above sea level with a 20% slope. The soil in the area is classified as Argissolo Vermelho, characterized by acidic pH (4.6) and low calcium and magnesium contents (1.0 and 0.5–0.6 mmolc.dm^-3, respectively), low available phosphorus (1–2 mg.dm^-3), and high aluminum saturation (28–36%). Potassium levels were moderate (43–83 mg.dm^-3), whereas manganese was elevated (29.4–41.4 mg.dm^-3). The soil also exhibited marginal organic matter (OM) content (14.1–20.0 g.dm^-3) and low base saturation (30–38%), reflecting the naturally low fertility typical of tropical savanna soils.

Figure 1
Aerial view of the experimental area at the Instituto Federal Fluminense, Bom Jesus do Itabapoana Campus, Rio de Janeiro, Brazil.

Plant material and propagation

The orchard comprises 38 trees, planted in 2017 in 40 × 40 × 40 cm pits, spaced 4 × 5 m apart. Prior to planting, the area was occupied by eucalyptus stands and low natural vegetation, which was manually cleared. Planting pits were opened individually and filled with a mixture of soil and sand (2:1, v/v), supplemented with 250 g of single superphosphate per pit. After planting, mulch consisting of dried Paspalum notatum grass was applied to reduce soil moisture loss.

The soil in the experimental area is classified as Red-Yellow Argisol, according to Brazilian Agricultural Research Corporation (2013). The soil profile adjacent to the experimental area is characterized by an A horizon extending from 0- to 0.20-m depth, with mean values of 421 g.kg^-1 sand, 127 g.kg^-1 silt, and 452 g.kg^-1 clay. The B horizon extends to approximately 2 m depth. In the upper landscape position (hilltop), characteristics associated with Dystrophic Red Latosol become more evident, indicating a transition from Red-Yellow Argisol to Dystrophic Red Latosol. The soil exhibits moderate structure and relatively lower compaction, allowing adequate root development throughout the profile.

No additional fertilization or soil correction practices were carried out after planting, and the orchard has been maintained under minimal management conditions since establishment.

Seedlings were obtained via seed propagation in the same year, in accordance with Costa et al. (2022). Mechanical scarification involved cutting a slit in the hilum region, while chemical scarification followed the protocol described by Oliveira et al. (2006). Germination occurred in Vivatto^® vegetable substrate within a greenhouse providing 50% shading, at temperatures ranging from 25 to 40°C and relative humidity between 35 and 100%. Emergence reached 55% at 45 days after sowing (DAS), and the plants began juvenile-adult phase transition in spring 2021.

Plant biometric characterization

Plants were evaluated for biometric and qualitative traits, as well as fruit development. The biometric variables measured were height, stem diameter, and canopy diameter in the row and inter-row directions, recorded nnually from 2021 to 2025. Height was measured using a tape measure attached to a telescopic pole, and canopy diameter was obtained from two perpendicular measurements taken with a measuring tape and poles positioned at the canopy edges, allowing calculation of the mean diameter based on the largest canopy dimensions. Stem diameter was measured at 0.10 m above the stem base to account for the multistemmed growth observed in some individuals.

Data on height, stem diameter, and canopy diameter were analyzed using R software. Analysis of variance (ANOVA) was applied at a 5% significance level (p < 0.05). When significant differences were detected via the F-test, means were compared using Duncan’s multiple range test at 5% probability to identify statistically distinct groups among the years evaluated.

Qualitative traits included the number of primary branches emerging directly from the trunk and the number of stems per individual, to identify occurrences of a multistemmed growth habit, a species-specific characteristic (Fig. 2). Trunks were characterized based on bark texture (smooth or rough), presence or absence of bark exfoliation, and coloration (grayish or brownish), while inflorescences were classified as burgundy or purplish green (Fig. 2). Scores were assigned to standardize assessments, as follows: 1 for smooth and 2 for rough bark; 1 for presence and 2 for absence of bark exfoliation; 1 for grayish and 2 for brownish coloration; and 1 for burgundy and 2 for purplish green inflorescences. Qualitative data were presented descriptively, without statistical testing.

Figure 2
Examples of (a) multistemmed growth habit with rough bark, (b); single smooth trunk, (c) burgundy inflorescence, and (d) purplish green inflorescence.

Phenological assessments

To characterize the orchard’s phenology, the relationships between defoliation and budburst and climatic variables were analyzed, along with the cumulative degree-days in each phenological phase and inflorescence monitoring.

Defoliation and budburst levels were monitored over a 14-month period (2023 to 2024) using a visual scoring system (Fig. 3). For defoliation, scores were assigned as follows:

Figure 3
Pequi trees at different defoliation levels: (a) plants that had lost up to 25% of their foliage; (b) 25 to 50%; (c) 50 to 75%; and (d) over 75%.

• 1: plants that had lost up to 25% of their foliage;

• 2: 25 to 50%;

• 3: 50 to 75%;

• 4: over 75%.

Defoliation and sprouting were assessed using the same visual scoring approach, with specific score ranges defined for each variable as follows:

• 0: up to 25%;

• 1: 25 to 50%;

• 2: 50 to 75%;

• 3: over 75%.

Daily and monthly meteorological data on rainfall and temperature were obtained from a Brazilian National Institute of Meteorology weather station in Itaperuna, Rio de Janeiro state. Statistical analyses were performed in R software. Relationships among defoliation, budburst, rainfall, and mean temperature were examined using Spearman’s rank correlation coefficient (ρ) for non-parametric variables, at a significance level of p < 0.05. The correlation matrix was graphically represented using the Corrplot package, showing both coefficients and correlation gradients.

The study of cumulative degree-days was based on identifying the main stages of vegetative and reproductive cycles, from floral bud differentiation to physiological maturation and natural fruit drop. A visual phenological scale was developed (Fig.4), comprising the following phases: closed floral bud, anthesis, fall of incomplete flowers, fruit set, fruit development, maturation, and fruit drop.

Figure 4
Scale of the phenological phases of pequi (Caryocar brasiliense) in northwestern Rio de Janeiro state, Brazil.

The duration of each phenophase was determined by calculating degree-days (DD), using Eq. 1:

(1) $\mathrm{DD}=\frac{\mathrm{TMAX}-\mathrm{TMIN}}{2}-\mathrm{Tb}$

where: TMAX: the daily maximum temperature (°C); TMIN: the daily minimum temperature (°C); Tb: the base temperature (13°C).

The cumulative DD was calculated by summing daily values across each subperiod, enabling estimation of the thermal requirements and duration of each developmental stage.

A total of 12 reproductive individuals were evaluated in 2023 and 16 in 2024, monitored periodically from June 2023 to December 2024, totaling 19 months of observations. During each visit, phenophases were recorded by direct canopy inspection. In addition, the number of inflorescences was counted on 12 adult trees in 2023. Both inflorescence and degree-day data were summarized using descriptive statistics.

RESULTS AND DISCUSSION

Plant vigor assessment and phenotypic characterization

Before presenting the biometric results, it is important to consider the environmental and edaphic conditions under which the orchard was established. The region presents a seasonal climatic pattern, with variations in rainfall and temperature throughout the year, which directly influence plant growth dynamics. Although periods of water deficit may occur, these conditions appear to have been sufficient to sustain plant development.

In addition, the soil is characterized by low natural fertility, with acidic pH, low levels of essential nutrients, and high aluminum saturation, which would typically limit plant growth. However, the presence of a deep soil profile, associated with the transition between Argissolo and Latossolo, may have favored root development and water access. Combined with basic practices adopted at planting and the absence of further fertilization, these conditions highlight the ability of C. brasiliense to establish and develop under low-input and potentially restrictive environments.

Analysis of variance indicated a significant effect of year on the stem diameter of pequi trees (F = 7.74; p < 0.001) (Fig. 5), revealing substantial variation across the evaluated years. Duncan’s multiple comparison test showed that the highest diameter (24.20 cm) was recorded in 2025, differing statistically from the other years. Intermediate values were obtained in 2022 and 2023 (19.20 and 19.04 cm, respectively), with no significant differences between these years, while 2024 exhibited values statistically similar to those observed in 2023 but higher than 2021. The lowest mean diameter (15.67 cm) was recorded in 2021, differing significantly from all other years. These results indicate a progressive increase in diameter growth over time, with superior performance in 2025, likely associated with favorable climatic conditions and the advancing maturity of the orchard.

Figure 5
Distribution of stem diameter (cm) from 2021 to 2025. Boxplots indicate the median (central line), quartiles (box), minimum and maximum values exceeding outliers (vertical lines), and individual outliers (red dots). Blue dots represent the annual mean.

Overall, the results indicated a consistent increase in stem diameter over the years, suggesting that the plants are still in an active growth phase. The higher values observed in 2025 may be related to more favorable environmental conditions during this period, in addition to the natural development of the orchard over time. This pattern indicates that C. brasiliense shows good growth performance under the studied conditions, even without intensive management.

ANOVA also revealed statistically significant differences in plant height over the evaluated years (F4,180 = 14.82; p < 0.001) (Fig. 6), indicating a relevant temporal effect on vegetative growth. Duncan’s multiple comparison test highlighted a clear pattern of progressive increase in mean plant height over the study period. The greatest mean height was recorded in 2025 (5.52 ± 1.39 cm), significantly higher than in 2024 (4.85 ± 1.27 cm), 2023 (4.66 ± 0.73 cm), and 2022 (4.65 ± 1.04 cm), which did not differ statistically. By contrast, plants in 2021 had the lowest mean height (3.61 ± 0.86 cm), differing statistically from all other years. This pattern suggests continuous and cumulative growth over time, potentially influenced by environmental factors, ecological adaptation, and species-specific biological traits. The results underscore the importance of considering temporal effects in plant development studies, since significant changes can occur even over relatively short annual intervals, reflecting both species plasticity and environmental influences across years.

Figure 6
Distribution of plant height from 2021 to 2025. Boxplots represent the median (central line), quartiles (box), and top and bottom horizontal lines, with red dots indicating outliers. Blue circles represent the mean height for each year. Different letters above the boxes indicate statistically significant differences among years, according to Duncan’s multiple comparison test (α = 0.05).

Overall, the results showed a clear increase in plant height over the years, indicating continuous and consistent development of the trees. The higher values observed in the most recent years suggest that the plants are well established and responding positively to the growing conditions. This pattern reinforces the good performance of C. brasiliense in the study area, even under the environmental and soil limitations described. This response may also be related to the depth of the soil profile, which likely favors root penetration and access to water in deeper layers.

ANOVA also revealed statistically significant differences between years for inter-row (INTR) and in-row canopy diameter (INR) (Fig. 7). For INTR, the effect of year was significant (F = 7.384; p < 0.001), indicating variation in mean values across 2023, 2024, and 2025. Similarly, for INR, ANOVA showed significant differences between years (F = 5.529; p = 0.005). Duncan’s multiple comparison test identified the specific years that differed. Mean INTR was significantly higher in 2025 (6.03 m, group “a”) when compared to 2023 (4.98 m) and 2024 (5.37 m), which did not differ statistically (group “b”). Mean values for INR were also significantly higher in 2025 (6.14 m, group “a”) than in 2023 (5.36 m) and 2024 (5.19 m), both classified in group “b.” These results indicate a trend of increasing values for both parameters over the period evaluated, suggesting consistent changes in conditions or characteristics associated with canopy diameter across the years.

Figure 7
Distribution of in-row canopy diameter (INR) and inter-row (INTR) values (m) across 2023, 2024, and 2025. Boxes represent the interquartile range (Q1–Q3), the internal horizontal line indicates the median, and external points represent outliers. Duncan’s multiple comparison test is indicated by letters above the boxes: years with different letters denote significantly different means at a 5% significance level (p < 0.05).

Ferreira et al. (2015) described C. brasiliense as a large tree with a shrubby to arboreal growth form, stem diameters ranging from 0.27 to 0.51 m, and heights between 5.91 and 10.11 m. Other studies also report that the species can reach considerable height and stem diameter under natural conditions (Leite et al. 2006). Although these descriptions refer to naturally occurring individuals with unknown age and growth history, the results of the present study show that cultivated trees in northwestern Rio de Janeiro state, even at a relatively young stage, already display biometric values comparable to those reported under natural conditions. This pattern suggests that the observed growth is associated with favorable environmental conditions and the inherent plasticity of the species, even in the absence of fertilization, pruning, or other management practices after establishment.

In recent years, Brazilian fruit production has increasingly adopted more intensive systems, particularly through orchard densification, aiming to optimize land use and increase productivity (Fremout et al. 2022). In this context, the favorable biometric performance observed in the present study indicates that C. brasiliense has potential for inclusion in such systems, if factors such as spacing and canopy management are properly adjusted. The use of denser planting arrangements may therefore represent a promising alternative, combining productivity gains with the sustainable use of a native species of ecological and socioeconomic relevance.

The predominance of grayish stem coloration (71.05%) indicates relative uniformity for this trait, possibly influenced by both genetic and environmental factors. Smooth bark, observed in more than half of the individuals (55.26%), may function as a potential defense against insects by reducing their ability to adhere to the stem (Ferrenberg and Mitton 2014). Bark exfoliation, present in 78.95% of the trees, is a common feature among Cerrado species and may be associated with adaptation to high-light environments and protection against pathogens and fire (Pellegrini et al. 2016, Loram-Lourenço et al. 2020, Chiminazzo et al. 2023).

The predominance of purplish green inflorescences (91.67%) is consistent with previous descriptions of the species (Ferreira et al. 2015). Although most trees (81.58%) exhibited a single stem, the occurrence of multistemmed individuals (18.42%) highlights the morphological variability of the species. The average number of 2.8 primary branches reflects the natural branching pattern of pequi trees, which is relevant for light interception, canopy structure, and management practices in cultivated systems (Pellegrini et al. 2016, Fremout et al. 2022).

Phenological assessments

Caryocar brasiliense is typically associated with regions characterized by seasonal rainfall and warm temperatures, conditions commonly found in the Brazilian Cerrado (Leite et al. 2006, Fagundes et al. 2007). In these environments, the species develops under a marked dry and wet season, demonstrating adaptation to water limitation and high temperatures (Mota et al. 2025). In 2023, the study site received 883.2 mm of rainfall, with mean temperatures ranging from 19.97 to 31.08°C. Despite being below the optimal range reported for the species, rainfall was close to recommended values (Leite et al. 2006). Although temperatures in Itaperuna, near the orchard in Bom Jesus do Itabapoana, RJ (Fig. 8), were elevated compared to natural occurrence areas, the trees established successfully and began fruit production eight years after planting.

Figure 8
Monthly mean, minimum, and maximum temperatures and rainfall between 2023 and 2025.

Spearman’s correlation analysis (Fig. 9) showed that budburst was positively associated with climatic factors, particularly mean temperature (ρ = 0.34) and rainfall (ρ = 0.25), indicating greater vegetative activity during warmer and wetter periods. This type of response is commonly observed in tropical species, in which growth is closely linked to seasonal climatic variation (Alberton et al. 2023, Macphie and Phillimore 2024). Climatic variables themselves were highly correlated (ρ = 0.94), highlighting the strong coupling between temperature and rainfall in the region.

Figure 9
Spearman correlation matrix (ρ) between phenological variables (defoliation and budburst) and climatic variables (rainfall and mean temperature) in a pequi orchard in Bom Jesus do Itabapoana, RJ, Brazil, from 2023 to 2024. Values indicate correlation coefficients between variable pairs. Blue shades represent positive correlations and red shades negative correlations, while color intensity indicates coefficient magnitude.

In contrast, defoliation showed no significant correlation with either rainfall or temperature, with low coefficients (ρ = -0.11 and ρ = -0.10, respectively). Similarly, the correlation between defoliation and budburst was low (ρ = 0.01), indicating little dependency between these processes. This suggests that leaf shedding may be less directly influenced by short-term climatic variation, as also reported for Cerrado species (Mota et al. 2025).

A possible explanation for this behavior is that trees growing on hilltops develop deeper root systems, allowing access to water stored in deeper soil layers. Under these conditions, defoliation may be less dependent on rainfall, whereas budburst responds more directly to favorable environmental conditions.

The phenological pattern observed in this study does not strictly follow a fixed sequence of complete defoliation followed by budburst and flowering. Instead, these events may overlap or vary depending on local conditions, indicating a certain flexibility in the phenological behavior of C. brasiliense. Similar variability has been described in previous studies, which report that phenological events in this species are influenced by climatic seasonality but do not always occur in a rigid sequence (Leite et al. 2006, Fagundes et al. 2007).

However, variations in this pattern may occur. The occurrence of flowering in individuals that still retain leaves, as observed in this study, has also been reported in the literature and is often associated with local environmental conditions or differences among individuals (Leite et al. 2006, Silva and Tubaldini 2013).

Although not included in the formal experimental design, additional observations made during a subsequent cycle indicated more homogeneous phenological behavior in the orchard, associated with lower temperatures at the end of winter. Cold conditions appeared to act as a stimulus for both defoliation and flowering, synchronizing these events among individuals. By the first week of September, most trees were fully defoliated, while flowering commenced uniformly from mid-August. These observations reinforce the role of temperature variation as a potential trigger for intensifying defoliation and flowering. Defoliation may also contribute to plant health by removing senescent or diseased tissues and increasing litter input to the soil. Although these observations were not subjected to statistical analysis, they are consistent with previous studies highlighting the influence of climatic variation on pequi leaf fall and flowering (Leite et al. 2006, Fagundes et al. 2007).

Overall, the positive relationship between budburst and climatic factors reinforces the influence of temperature and rainfall on vegetative activity. The strong correlation between these variables highlights their combined role in regulating phenological responses, which is essential for understanding the dynamics of C. brasiliense under the studied conditions (Leite et al. 2006, Fagundes et al. 2007).

Beyond the correlation between budburst and defoliation with climate, thermal accumulation expressed as degree-days represents a key variable for understanding phenological development. It allows the estimation of the time required for transitions between reproductive stages, since temperature directly influences developmental rates in plants (Colodetti et al. 2025). In this sense, analyzing phenological stages based on thermal accumulation provides a more accurate understanding of the timing of reproductive events.

The reproductive phase of pequi followed the seasonal pattern of temperature and rainfall recorded between January 2023 and November 2024 (Fig. 8). Anthesis, which required 217.08 degree-days over 20 days (Table 1), occurred under slightly higher temperatures and increasing humidity associated with the onset of rainfall. This suggests that short-term increases in temperature combined with initial water availability may favor the transition to reproductive stages, as observed in other tropical species (Alberton et al. 2023).

The fall of incomplete flowers occurred over a short interval (9.5 days) and required lower thermal accumulation (105.8 degree-days) (Table 1), indicating a rapid adjustment phase following flowering, possibly related to resource allocation and environmental filtering of reproductive structures.

Table 1
Cumulative degree-days and mean duration in days for each phenological stage of pequi in the 2023/2024 and 2024/2025 cycles, in northwestern Rio de Janeiro state, Brazil.

Stage	23/24	24/25	Mean	23/24	24/25	Mean
	DD			DUR
CFB	3,088.67	2,974.78	3,031.70	237.00	212.00	224.50
A	126.65	307.50	217.07	12.00	28.00	20.00
FIF	138.50	73.10	105.80	12.00	7.00	9.50
FS	290.20	434.90	362.55	21.00	37.00	29.00
FD	163.25	439.35	301.30	12.00	32.00	22.00
M	103.80	107.10	105.45	8.00	8.00	8.00
FDp	563.92	363.85	463.90	36.00	27.00	31.50
TOTAL	4,475.00	4,700.57	4,587.78	338.00	351.00	344.50

• DD: cumulative degree-days; DUR: mean duration in days; CFB: closed floral bud; A: anthesis; FIF: petal fall; FS: fruit set; FD: fruit development; M: maturation; FDp: fruit drop. Values represent the arithmetic mean across the two agricultural cycles assessed.

The fruit set phase required higher thermal accumulation (362.55 degree-days) (Table 1) and coincided with increased rainfall, suggesting that both temperature and water availability contribute to the stabilization of developing fruits. Fruit development, with 301.3 degree-days over 22 days, occurred under conditions of adequate moisture, favoring rapid growth. These results indicated that the progression of reproductive stages depends on the combined effects of thermal accumulation and water availability, a pattern commonly reported for tropical fruit species (Alberton et al. 2023, Mota et al. 2025).

The maturation phase, which required 105.45 degree-days over eight days (Table 1), occurred under gradually decreasing rainfall and moderate temperatures (~24°C), indicating a shorter and less thermally demanding phase. Physiological fruit drop, requiring 463.89 degree-days over 31.5 days, occurred during periods of increased rainfall, suggesting that environmental conditions may influence the completion of the reproductive cycle and fruit abscission processes.

Overall, the complete cycle of C. brasiliense, from anthesis to fruit drop, lasted 344 days, closely following the annual pattern of temperature and rainfall in the region. This result showed that thermal accumulation is an effective indicator of phenological progression and reinforces the strong dependence of the species on seasonal climatic conditions. Similar relationships between temperature, water availability, and phenological timing have been reported for other tropical and savanna species, highlighting the role of climate as a primary driver of reproductive dynamics (Alberton et al. 2023, Macphie and Phillimore 2024).

The relationship between phenological development and climatic conditions observed in this study reinforces the importance of water and temperature as primary drivers of plant dynamics in tropical ecosystems. In seasonal environments such as the Cerrado, plant phenology is strongly regulated by the interplay between water availability and climatic seasonality, which directly affects growth, reproduction, and ecosystem productivity (Alberton et al. 2023).

In the case of C. brasiliense, this dependence on climatic conditions is associated with adaptive strategies that allow the species to persist under water-limited environments. Previous studies have shown that the species exhibits traits such as deep root systems, underground reserve structures, and tolerance to drought and high temperatures, which support its establishment and development even under restrictive conditions (Mota et al. 2025). These characteristics help explain the phenological flexibility observed in the present study, particularly the capacity to maintain vegetative and reproductive activity under variable environmental conditions.

The phenological pattern identified, with flowering during the dry season and fruiting during the rainy season, is consistent with previous studies conducted in the Cerrado, indicating a strong synchronization between reproductive events and seasonal climatic conditions (Leite et al. 2006). However, the variability observed among individuals and across periods suggests that this pattern is not rigid, but rather responsive to local environmental conditions.

Phenological responses are not only important for describing plant development but also play a central role in regulating ecosystem processes and productivity. In tropical environments, the timing of vegetative and reproductive events is strongly controlled by the interaction between water availability and climatic conditions, which directly affects plant performance (Alberton et al. 2023). Moreover, phenology is highly sensitive to environmental variation, and shifts in temperature and precipitation patterns can alter the timing and synchronization of biological events, with potential consequences for plant reproduction and species interactions (Macphie and Phillimore 2024).

In productive terms, climatic variability can also influence fruit yield, as demonstrated for Caryocar species, in which precipitation has been identified as a key factor regulating fruit production (Gomes et al. 2022). These findings indicate that the phenological patterns observed in the present study are not only ecologically relevant, but also directly linked to productivity and resource availability.

In this context, changes in rainfall distribution and temperature regimes may alter the timing and synchronization of phenological events. Since phenology is closely linked to reproductive success, such changes may affect both natural populations and production systems. Therefore, understanding the phenological responses of C. brasiliense becomes essential not only for ecological studies, but also for its domestication, management, and use in restoration and sustainable production systems.

The results also highlighted the sensitivity of C. brasiliense phenology to climatic conditions, particularly temperature and precipitation, indicating that ongoing climate change may influence the timing and duration of its phenological cycle. As a species native to the Brazilian Cerrado, one of the most threatened and degraded biomes, C. brasiliense has evolved under conditions of high temperature and seasonal water limitation. Its deep root system allows access to water in deeper soil layers, reducing sensitivity to short-term water deficits and contributing to its persistence in drought-prone environments.

In the context of climate change, with rising temperatures and increasingly irregular and reduced precipitation in many regions, environmental conditions similar to those of the Cerrado are expected to expand. Therefore, understanding its phenological responses under current environmental conditions is essential not only for predicting future changes in its reproductive cycle and fruit production, but also for identifying its potential as a resilient species for use in regions increasingly affected by climatic constraints.

The results of this study highlighted the potential of C. brasiliense for domestication and cultivation under low-input conditions. The species showed satisfactory and consistent phenological performance even in low-fertility soils and without management, indicating strong adaptability. This phenological flexibility suggests that C. brasiliense can adjust to local environmental conditions, supporting its use in sustainable production systems, such as agroforestry and orchards aimed at restoration, since it does not require intensive management.

In the region where the orchard was established, flowering began in the first week of August 2023 and peaked shortly thereafter (Fig. 10). New inflorescences continued to emerge until the third week of November, resulting in an average duration of 110 days. Fruiting began in the first week of September and extended until the end of December, lasting approximately three uninterrupted months (Fig. 10). Complete synchrony among individuals was not observed, with some trees flowering earlier and others later.

Figure 10
Flowering and fruit harvest timing and intensity of pequi during the 2023 season in northwestern Rio de Janeiro, Brazil.

Peak flowering occurred outside the usual period observed in major production regions but coincided with the onset of production in the Federal District and Mato Grosso do Sul state, Brazil. The fruiting period, starting in September and peaking in early December, aligns with those of the Federal District, Mato Grosso do Sul, and São Paulo, Brazil.

Phenological studies in different regions of Brazil indicate variability in the vegetative cycle of C. brasiliense. In seasonal environments, leaf flush and defoliation are commonly observed during the dry season, often occurring under water deficit conditions and with variable synchrony among individuals (Leite et al. 2006). Fagundes et al. (2007) studied areas in northern Minas Gerais dominated by pastures, regenerating Cerrado, and Cerradão, and reported that defoliation occurs year-round, with peaks in August-September and February-March, following a biennial pattern.

The reproductive phase of C. brasiliense typically begins with floral bud emergence in the late dry season, with flowering extending from August to October, depending on local environmental conditions (Leite et al. 2006). Variations in the timing and duration of flowering have been reported across regions, reflecting the influence of climatic factors such as temperature and water availability. These observations are consistent with the findings of Pirani et al. (2009), who reported the highest flowering peak in the same period. In northwestern Rio de Janeiro, under cultivation from seedlings, the first trees flowered 61 months after planting, with floral clusters observed until October (Costa et al. 2022). Seed-propagated trees exhibit high phenological variability, potentially affecting both the timing of fruiting and fruit quality.

CONCLUSION

This study demonstrates that C. brasiliense exhibits adaptive potential to tropical savanna climates, with consistent reproductive cycles and well-defined phenological stages. Biometric analyses indicate that the orchard is still developing, since the variables obtained in 2025 exceeded those of the previous years. In addition to growth variability, the population showed notable morphological diversity, including differences in inflorescence color, bark texture, trunk roughness, and degrees of polycauly, suggesting high intra-populational variability that may confer ecological resilience. Phenological characterization showed that the reproductive cycle aligns closely with local rainfall and temperature patterns, with floral emergence dependent on cumulative degree-days and fruiting synchronized with rainfall, enabling harvest in December. These results supply valuable phenological and growth benchmarks that can support future research, conservation programs, and sustainable pequi cultivation, reinforcing its potential for both productive use and ecological restoration.

ACKNOWLEDGMENTS

The authors thank the Graduate Program in Plant Production and the Plant Science Laboratory for providing the facilities and resources necessary to conduct the analyses, as well as the laboratory team for their assistance during the experimental activities. Special thanks are extended to the Instituto Federal Fluminense, Bom Jesus do Itabapoana Campus, for granting access to the orchard and providing logistical support during fieldwork, which were fundamental to the development of this study.

• How to cite:
  Silva, R. M. R., Marinho, C. S., Galvão, S. P., Santos, R. F., Silva, J. E. V. C., Silva, A. E. and Costa, E. S. (2026). Phenotypic variability and phenological behavior of Caryocar brasiliense under tropical savanna climate conditions. Bragantia, 85, e20250261. https://doi.org/10.1590/1678-4499.20250261
• FUNDING
  Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro
  Grant No: E-26/200.674/2022 (274926)
• DECLARATION OF USE OF ARTIFICIAL INTELLIGENCE TOOLS
  The authors declare that no artificial intelligence tools were used in the development, writing, or revision of this manuscript.

DATA AVAILABILITY STATEMENT

All data generated or analyzed during this study are included in this article.

REFERENCES

• Afonso, S. R. (2022). Innovation perspectives for the bioeconomy of non-timber forest products in Brazil. Forests, 13, 2046. https://doi.org/10.3390/f13122046
  » https://doi.org/10.3390/f13122046
• Alberton, B., Martin, T. C. M., Rocha, H. R., Richardson, A. D., Moura, M. S. B., Torres, R. S. and Morellato, L. P. C. (2023). Relationship between tropical leaf phenology and ecosystem productivity using phenocameras. Frontiers in Environmental Science, 11, 1223219. https://doi.org/10.3389/fenvs.2023.1223219
  » https://doi.org/10.3389/fenvs.2023.1223219
• Chiminazzo, M., Bombo, A., Charles-Dominique, T. and Fidelis, A. (2023). Bark production of generalist and specialist species in savannas and forests of the Cerrado. Annals of Botany, 131, 613-621. https://doi.org/10.1093/aob/mcad014
  » https://doi.org/10.1093/aob/mcad014
• Colodetti, T. V., Tomaz, M. A., Sartori, L., Salles, R. A., Souza, I. V., Senra, J. F. B. and Rodrigues, W. N. (2025). From anthesis to harvest: evolution of the fructification of Conilon coffee genotypes grown at a transitional altitude. Bragantia, 84, e20250028. https://doi.org/10.1590/1678-4499.20250028
  » https://doi.org/10.1590/1678-4499.20250028
• Costa, E. S., Marinho, C. S., Amaral, C. O. and Souza, J. A. G. (2022). Early germination of pequi seeds and post-planting seedling performance in the northwestern of Rio de Janeiro. Floresta e Ambiente, 29, e20210099. https://doi.org/10.1590/2179-8087-FLORAM-2021-0099
  » https://doi.org/10.1590/2179-8087-FLORAM-2021-0099
• Cunha, N. R. S., Lima, J. E., Gomes, M. F. M. and Braga, M. (2008). The intensity of agricultural exploitation as an indicator of environmental manipulation in the Cerrado region, Brazil. Revista de Economia e Sociologia Rural, 46, 291-323. https://doi.org/10.1590/S0103-20032008000200002
  » https://doi.org/10.1590/S0103-20032008000200002
• [EMBRAPA] Brazilian Agricultural Research Corporation (2013). Brazilian soil classification system. 3rd ed. Brasília: EMBRAPA.
• Fagundes, M., Araújo, L. S. and Nunes, Y. R. F. (2007). Effects of habitat successional stage on the phenology of pequi tree (Caryocar brasiliense: Caryocaraceae). Revista Brasileira de Biociências, 5, 144-146.
• Ferreira, G. A., Naves, R. V., Chaves, L. J., Veloso, V. R. S. and Souza, E. R. B. (2015). Fruit production of natural populations of pequi in the state of Goiás. Revista Brasileira de Fruticultura, 37, 121-129. https://doi.org/10.1590/0100-2945-404/13
  » https://doi.org/10.1590/0100-2945-404/13
• Ferrenberg, S. and Mitton, J. (2014). Smooth bark surfaces may defend trees against insect attacks: reviving a slippery hypothesis. Functional Ecology, 28, 837-845. https://doi.org/10.1111/1365-2435.12228
  » https://doi.org/10.1111/1365-2435.12228
• Figueiredo, R. W., Maia, G. A. and Figueiredo, E. A. (1989). Propriedades físico-químicas e composição dos ácidos graxos da fração lipídica da polpa e amêndoa do pequi (Caryocar coriaceum Wittm). Ciência Agronômica, 20, 135-139.
• Fremout, T., Thomas, E., Taedoumg, H., Briers, S., Gutiérrez-Miranda, C., Alcázar-Caicedo, C., Lindau, A., Kpoumie, H., Vinceti, B., Kettle, C., Ekué, M., Atkinson, R., Jalonen, R., Gaisberger, H., Elliott, S., Brechbühler, E., Ceccarelli, V., Krishnan, S., Vacik, V., Wiederkehr-Guerra, G., Salgado-Negret, B., González, A. A., Ramírez, W., Moscoso-Higuita, L. G., Vásquez, A., Cerrón, J., Maycock, C. and Muys, B. (2022). Diversity for restoration (D4R): guiding the selection of tree species and seed sources for climate-resilient restoration of tropical forest landscapes. Journal of Applied Ecology, 59, 664-679. https://doi.org/10.1111/1365-2664.14079
  » https://doi.org/10.1111/1365-2664.14079
• Gomes, B. H., Faria, M. V., Mendes, M. G., Bonetti, A. M., Oliveira Júnior, R. J. and Nogueira, A. P. O. (2022). Genetic diversity and correlation between morphological traits of pequi fruits (Caryocar brasiliense Camb.) with and without thorns at the endocarp. Anais da Academia Brasileira de Ciências, 94, e20210016. https://doi.org/10.1590/0001-3765202220210016
  » https://doi.org/10.1590/0001-3765202220210016
• Leite, G. L. D., Veloso, R. V. S., Zanuncio, J. C., Fernandes, L. A. and Almeida, C. I. M. (2006). Phenology of Caryocar brasiliense in the Brazilian Cerrado region. Forest Ecology and Management, 236, 286-294. https://doi.org/10.1016/j.foreco.2006.09.013
  » https://doi.org/10.1016/j.foreco.2006.09.013
• Loram-Lourenço, L., Farnese, F., Sousa, L., Alves, R., Andrade, M., Almeida, S., Moura, L., Costa, A., Silva, F., Galmés, J., Cochard, H., Franco, A. C. and Menezes-Silva, P. E. (2020). A structure shaped by fire, but also by water: ecological consequences of variability in bark properties in 31 species of the Brazilian Cerrado. Frontiers in Plant Science, 10, 1718. https://doi.org/10.3389/fpls.2019.01718
  » https://doi.org/10.3389/fpls.2019.01718
• Macphie, K. H. and Phillimore, A. B. (2024). Phenology. Current Biology, 34, R183-R188. https://doi.org/10.1016/j.cub.2024.01.007
  » https://doi.org/10.1016/j.cub.2024.01.007
• Mota, V. O., Ribeiro, L. M., Figueiredo, M. G. F., Mercadante-Simões, M. O., Mota, M. F. C., Trindade, D. A. R. and Lopes, P. S. N. (2025). Impact of water availability on Caryocar brasiliense (Caryocaraceae) seedling establishment. Journal of Arid Environments, 229, 105379. https://doi.org/10.1016/j.jaridenv.2025.105379
  » https://doi.org/10.1016/j.jaridenv.2025.105379
• Oliveira, R. P., Scivittaro, W. B. and Radmann, E. B. (2006). Seedcoat chemist scarification to optimize the emergency and growth of Trifoliata rootstock Pesquisa Agropecuária Brasileira, 41, 1429-1433. https://doi.org/10.1590/S0100-204X2006000900012
  » https://doi.org/10.1590/S0100-204X2006000900012
• Pellegrini, I. A., Franco, A. C. and Hoffmann, W. A. (2016). Changes in functional traits increase the risk of tree mortality caused by fire in savanna and tropical forest biomes. Global Change Biology, 22, 1235-1243. https://doi.org/10.1111/gcb.13110
  » https://doi.org/10.1111/gcb.13110
• Pirani, F. R., Sanchez, M. and Pedroni, F. (2009). Phenology of a tree community in cerrado sensu stricto, Barra do Garças, MT, Brazil. Acta Botanica Brasilica, 23, 1096-1109. https://doi.org/10.1590/S0102-33062009000400019
  » https://doi.org/10.1590/S0102-33062009000400019
• Santos, P., Porto, A. G., Silva, F. S. and Furtado, G. F. (2010). Physicochemical and sensory evaluation of pequi (Caryocar brasiliense Camb.) subjected to dehydration. Revista Brasileira de Produtos Agroindustriais, 12, 115-123. https://doi.org/10.15871/1517-8595/rbpa.v12n2p115-123
  » https://doi.org/10.15871/1517-8595/rbpa.v12n2p115-123
• Silva, M. N. and Tubaldini, M. A. S. (2013). The gold of the Cerrado: the dynamics of pequi extraction in northern Minas Gerais. Revista Geoaraguaia, 3, 293-317.

Edited by

• Section Editor:
  Gabriel Constantino Blain https://orcid.org/0000-0001-8832-7734

Publication Dates

• Publication in this collection
  27 July 2026
• Date of issue
  2026

History

• Received
  28 Nov 2025
• Accepted
  27 Apr 2026