Abstract
Introduction:
Cacao plants can translocate heavy metals from the soil to vegetative tissues, potentially leading to concentrations that exceed maximum permissible limits. This study aimed to analyze the geospatial distribution of cadmium (Cd) in soil profiles of cacao (Theobroma cacao L.) plantations in the Amazon basin of Peru.
Methods:
The research was conducted in the localities of Jeberillos, Arahuante, and Luz del Oriente, within the department of Loreto. In each locality, 10 plots of 100 m2 were selected, totaling 30 plots. Cd concentrations and physicochemical soil parameters were evaluated.
Results:
In cacao soils from Luz del Oriente at 0–30 cm depth, Cd concentrations averaged 1.11 ± 0.26 mg.kg-1, representing the highest values; whereas the lowest average was recorded in Arahuante at the same depth, with 0.61 ± 0.15 mg.kg-1. Cd levels in Luz del Oriente cacao soils (0–30 cm) were above the average among all localities, while Jeberillos and Arahuante soils showed comparatively lower Cd levels.
Discussion:
Overall, Cd in cacao soils was found to be strongly associated with edaphic factors such as pH, texture, effective cation exchange capacity (CEC), and cation composition. The maps represented the spatial distribution of the metal and guided the identification of potential accumulation zones. These findings provide essential insights for guiding agricultural management decision-making, environmental monitoring, and metal contamination risk assessment in cacao-growing areas and other agrarian systems in the Peruvian Amazon basin.
1 Introduction
The high global demand for cocoa positions this crop as an agrifood product of great economic value, recognized for its superior sensory and nutritional qualities (). In Peru, cacao is extensively cultivated by more than 100,000 farmers across >180,000 ha, yielding approximately 74,712.48 t annually with an average market price of USD$ 7.56 per kg. Production is concentrated mainly in the regions of Tumbes, Piura, Amazonas, Cajamarca, San Martín, Ucayali, Huánuco, Pasco, Junín, and Cusco (). Peru is also among the world’s leading exporters of cocoa, a sector that directly sustains the livelihoods of millions of smallholder farmers worldwide (). In recent years, the expansion of cacao cultivation has accelerated, driven by the sharp increase in international prices, which has significantly improved cocoa farmers’ household incomes. Yields are estimated to average 840 kg·ha-1, with cocoa typically grown in both conventional and organic agroforestry systems () in Colombia, Ecuador and Peru. Nevertheless, one of the major concerns threatening the sustainability of this sector is the presence of cadmium (Cd) in Amazonian soils.
Cd not only poses serious risks to human health but also exerts toxic effects on key physiological processes involved in plant growth and development, significantly affecting seed germination rates, transpiration rates, chlorophyll content, and biomass accumulation (Zhang et al., 2024). In crop plants, Cd toxicity impairs nutrient and water uptake and translocation, enhances oxidative damage, disrupts metabolic pathways, and inhibits both morphological and physiological functions (). The toxic effects of Cd extend from early developmental stages, such as seed germination, to yield suppression at maturity. Plant physiological functions, including water interactions, mineral nutrient uptake, and photosynthesis, are particularly affected by Cd. Moreover, Cd exposure induces metabolic alterations, either by directly affecting enzymes and metabolites or by promoting the production of reactive oxygen species, which in turn trigger oxidative stress. Beyond its biological impacts, Cd is also a major environmental pollutant, mainly due to its widespread industrial use and persistent contamination of soils and water, becoming a common, recent issue and activities and industrial wastewater (). Consequently, heavy metals such as Cd can severely constrain plant growth, reduce crop yields, and compromise the nutritional quality of food produced (; ).
Research on strategies to reduce cadmium (Cd) accumulation in cocoa beans is currently limited by an incomplete understanding of Cd transfer pathways within the cocoa tree (). This study provides an analysis of total cadmium concentrations in soils in relation to key edaphic variables, including pH, cation exchange capacity (CEC), carbon content, magnesium, among other. This knowledge is expected to have a significant impact on cocoa development and traceability, as it may help prevent the establishment of new plantations in areas where soil Cd levels approach maximum permissible limits. Accordingly, total soil Cd concentration, soil pH, and leaf Cd content have been identified as the best predictors of Cd accumulation in cocoa beans (). Moreover, elucidating the dominant factors governing heavy metal accumulation and their spatial heterogeneity in soils is essential for the implementation of science-based environmental management protocols. Collectively, these findings clarify the key drivers and spatial distribution patterns of Cd in the study area, providing a scientific basis for the targeted identification of contamination sources and the design of precise control strategies (). In addition, total soil Cd content decreased with increasing soil depth, consistent with the findings of , who reported a decline in total soil Cd from 0.377 to 0.196 mg kg-1 with increasing soil depth.
Several studies have shown that cacao plants can translocate soil heavy metals to their vegetative tissues, often resulting in concentrations that exceed the maximum permissible limits (; ; ). This is particularly critical for cadmium, which not only impairs plant organ function but can also be lethal to humans and animals (). In Peru, Cd concentrations in soils under cacao cultivation vary across geographic regions. Reported maximum values include 0.50, 0.53, 1.46, 0.10, 0.82, and 0.86 mg·kg-1 in the departments of Tumbes, Piura, Amazonas, Junín, Huánuco, and Ucayali, respectively (). In contrast, Cd contents in San Martín remain below the permissible threshold of 1.4 mg·kg-1 established by the Ministry of the Environment (), with values of 0.39 mg·kg-1 detected in cacao agroforestry systems. These lower concentrations are likely associated with the higher tree density typical of agroforestry systems compared to monoculture systems or the presence of arbuscular mycorrhizae (Rhizoglomus cacao) found in the Peruvian jungle ().
The departments with the highest average cacao yields are Junín (1,192 kg·ha-1), followed by La Libertad (1,161 kg·ha-1), Tumbes (1,039 kg·ha-1), Loreto (989 kg·ha-1), and San Martín (950 kg·ha-1) (). Although Loreto shows considerable potential for cacao production, no studies have yet been conducted to determine Cd concentrations in its soils. In this context, it is particularly relevant to analyze the geospatial distribution of Cd in cacao -cultivated soils of the Amazon basin in Peru, with special attention to edaphic zones and soil profile characteristics.
2 Materials and methods
2.1 Study area
The study was conducted in three localities within the Loreto department between May and December 2024. Loreto covers an area of 368,852 km2 and is characterized by a humid tropical climate, with minimal seasonal temperature variation (average 27 °C; range 20 °C–33 °C), high annual rainfall (2,500–3,000 mm), and consistently elevated relative humidity (87%–90%) throughout the year (). The three study sites included Arahuante (Arhant), located in the district of Lagunas, and Jeberillos (Jbrllos) and Luz del Oriente (Lhze), both situated in the district of Yurimaguas; all belong to the province of Alto Amazonas (Figure 1). The districts of Lagunas and Yurimaguas share the typical topography of the Amazonian lowlands, with flat to gently undulating relief and elevations ranging from 80 to 180 m.a.s.l. In Lagunas, the landscape is dominated by low terraces, tahuampas (seasonally flooded forests), and restingas (elevated levees), while Yurimaguas is characterized by medium terraces, fluvial deposits, and ancient river channels disconnected from the main river, which form seasonal lagoons. Both areas are underlain by poorly drained soils prone to temporary waterlogging (; ).
FIGURE 1
2.2 Sampling design
In this study, 10 plots of 100 m2 (10 m × 10 m) were established in each locality, for a total of 30 plots, following the methodology of Yu et al. (2019). Soil sampling was conducted by excavating pits measuring 1 × 1 m. Prior to excavation, the sites were cleared of weeds and litter, and samples were collected using a shovel and a metal bar. One sample was taken from each soil profile at depths of 0–30 cm and 30–60 cm, as several previous studies (; ; ) have reported significant variations in Cd stocks within similar soil depths. In total, 60 soil samples were obtained across the two soil profiles for subsequent analyses of cadmium concentrations and physicochemical soil properties. The soils in Yurimaguas (Alto Amazonas, Loreto) are predominantly alluvial, ranging from sandy loam to clayey textures, and belong mainly to the Ultisol order (). In other cases, Vertisols are present, such as those found in Luz del Oriente, which exhibit well-defined physical and chemical characteristics and a reddish-yellow coloration. Due to intensive productive pressure and management practices, many areas are severely degraded.
2.3 Determination of soil edaphic and nutritional parameters
Soil samples were collected using a shovel from different points within each plot, including the four corners and the center, and then homogenized to form a composite sample weighing approximately 1 kg. The samples were analyzed at the accredited laboratory of the National Institute of Agrarian Innovation (INIA), Tarapoto, Peru. Soil texture was determined using the hydrometer texture method; pH was measured potentiometrically in a 1:2.5 soil-to-water suspension following ; and soil organic matter (SOM) was quantified following the method.
Aluminum was measured by atomic absorption spectrophotometry using an air–nitrous oxide flame. Calcium and magnesium were determined following the Mexican Official Standard NOM-021-RECNAT-2000. Cation exchange capacity (CEC) was determined according to the Manual de Procedimientos de los Análisis de Suelos y Aguas con Fines de Riego ().
Available potassium (K) according to .
2.4 Determination of cadmium concentrations in soils
For Cd analysis, soil samples were collected, one from each subplot, following the Technical Sampling Guide developed by the Ministry of the Environment of Peru (). Total Cd in soil was determined using the EPA 3051A method (), which employs microwave-assisted acid digestion with nitric acid (HNO3) to simulate conventional heating-based extraction. After digestion, the concentration of Cd of soil was measured by flame atomic absorption spectrophotometry (direct aspiration) following EPA 700B:2007, using a standard solution prepared by dissolving 1 g of metallic cadmium in 20 mL of HNO3 (1:1) and diluting to 1 L with deionized water. The method detection limit was 0.2 mg·kg-1 ().
The Cd analysis was performed using inductively coupled plasma mass spectrometry (ICP–MS), following the quality assurance and quality control (QA/QC) procedures established in the US EPA methods and . Each analytical batch included a method blank to check for potential contamination, a fortified duplicate sample (5%) to assess precision and bias, and a laboratory control sample (LCS) with known concentrations to verify accuracy. Analytical performance was validated using standard reference materials (SRMs) from NIST, ensuring the reliability and traceability of the results.
The balances used in the preparation of this CRM are regularly calibrated with traceability to NIST standards. All volumetric dilutions are performed using calibrated Class A volumetric glassware. The certified concentration and its associated uncertainty were determined using the “high-performance ICP-OES” protocol developed by NIST, and both the certified concentration and uncertainty values are traceable to NIST SRM 3108, lot No. 130116. The uncertainty associated with the certified concentration represents the expanded uncertainty at a 95% confidence level, using a coverage factor of k = 2.
2.5 Statistical analyses
Statistical analyses were performed using the R software version 4.3.2 () through specialized packages such as ggplot2, corrplot, GGally, pheatmap, dplyr, readxl, and RColorBrewer (), which facilitated the processing, visualization, and comprehensive exploration of soil data collected from three localities in the Loreto department, Peru. A principal component analysis (PCA) was conducted with the factoextra package to reduce dimensionality and assess the joint variation of soil physicochemical variables. The results were displayed in a biplot, which enabled the visualization of the distribution of soil physicochemical parameters by locality and soil profile (0–30 and 30–60 cm), as well as the identification of variables that best explain the observed variability in the dataset. Subsequently, a geostatistical analysis of soil Cd was carried out using the Smart Map plugin in the QGIS software (version 3.34.14–Prizren) to fit theoretical variogram models, apply ordinary kriging interpolation, and evaluate the spatial dependence of the data. This approach supported the generation of thematic maps and the identification of Cd accumulation in soils. A cross-correlation matrix based on Pearson’s coefficient (p < 0.05) was then generated and visualized with the corrplot package to identify significant associations among variables. In addition, a heatmap was generated through the pheatmap package to examine clustering patterns among samples, while scatterplot matrices (GGgally package) were employed to explore correlations between numerical variables, color-coded by locality. Bar plots were also used to illustrate Cd distribution by locality and soil texture, aiding the interpretation of visual trends. Finally, analysis of variance (ANOVA) followed by Tukey’s post hoc test (p < 0.05) was applied to determine statistically significant differences among the mean values of soil physicochemical variables according to soil profile, thus providing inferential validation of the observed patterns.
3 Results and discussion
3.1 Physicochemical properties of soils across profiles in three localities
The highest cadmium content (Table 1) was recorded in cacao soils from Lhze at 0–30 cm depth, with an average of 1.11 ± 0.26 mg·kg-1, followed by Arhant soils at 30–60 cm depth with an average of 0.81 ± 0.11 mg·kg-1. No significant differences were observed between these two localities or soil profiles. The lowest Cd concentration was found in Arhant at 0–30 cm depth, averaging 0.61 ± 0.15 mg·kg-1. Soil pH varied across localities: Arhant showed an average of 3.58 ± 1.25 at 0–30 cm depth, along with a high sand content averaging 46.89% ± 7.29%. In contrast, Lhze presented the highest pH values, with an average of 6.30 ± 0.35. Regarding exchangeable cations, including potassium (K+), calcium (Ca2+), and magnesium (Mg2+), the highest values were observed in Arhant soils at 0–30 cm depth.
TABLE 1
| Physicochemical parameters | Tukey test by locality | Analysis of variance | ||||||
|---|---|---|---|---|---|---|---|---|
| Jbrllos | Arhant | Lhze | F | P-value | ||||
| 0–30 | 30–60 | 0–30 | 30–60 | 0–30 | 30–60 | |||
| pH | 5.55 ± 1.21 ab | 5.11 ± 1.16 b | 3.58 ± 1.25 c | 4.13 ± 0.21 c | 6.30 ± 0.35 a | 6.04 ± 1.24 a | 61.76 | <0.0001*** |
| K | 83.43 ± 11.41 b | 78.91 ± 13.23 bc | 41.44 ± 14.25 cd | 8.88 ± 5.88 d | 164.50 ± 32.73 a | 145.25 ± 32.52 a | 96.54 | <0.0001*** |
| CEC | 16.93 ± 4.05 bc | 20.48 ± 5.07 ab | 5.56 ± 1.65 cd | 3.62 ± 1.05 d | 28.99 ± 6.57 a | 29.26 ± 8.65 a | 37.8 | <0.0001*** |
| Cd | 0.77 ± 0.21 b | 0.73 ± 0.32 b | 0.61 ± 0.15 b | 0.81 ± 0.11 b | 1.11 ± 0.26 a | 0.75 ± 0.25 b | 9.31 | 0.0003*** |
| Sand | 47.74 ± 11.25 a | 41.32 ± 7.50 ab | 46.89 ± 7.29 a | 47.69 ± 8.96 a | 31.45 ± 7.60 bc | 23.63 ± 6.69 c | 26.85 | <0.0001*** |
| Clay | 24.62 ± 5.29 d | 34.94 ± 8.38 bc | 25.92 ± 6.32 cd | 36.65 ± 9.29 b | 40.16 ± 8.68 b | 49.48 ± 10.23 a | 32.04 | <0.0001*** |
| Ca | 14.78 ± 3.86 a | 13.87 ± 4.79 a | 0.86 ± 0.90 b | 0.44 ± 0.6 b | 24.72 ± 5.23 a | 24.31 ± 6.90 a | 35.05 | <0.0001*** |
| Mg | 1.60 ± 0.52 b | 1.13 ± 0.41 bc | 0.3 ± 0.2 cd | 0.12 ± 0.2 d | 3.64 ± 0.81 a | 3.26 ± 0.72 a | 95.92 | <0.0001*** |
Analysis of variance and Tukey’s post hoc test of soil physicochemical parameters across two soil profiles in three localities of the Loreto region.
Different letters within each physicochemical parameter indicate statistically significant differences among locations at p < 0.05, as determined by Tukey’s test. ns: not significant; *statistically significant; ***highly statistically significant.
Cacao soils in Arhant at 0–30 cm depth showed the lowest Cd content, characterized by low cation exchange capacity (CEC), high sand content, and low levels of Ca and Mg. These findings provide a basis for mitigation strategies through the application of soil amendments aimed at modifying soil properties and reducing Cd bioavailability in soil (Zhuang et al., 2023). In Lhze, the average soil pH was 6.17, while and emphasized that soil pH plays a critical role in regulating Cd bioavailability, noting that Cd availability in soil increases markedly when pH falls below 6.5. The elevated Cd content observed in Lhze at 0–30 cm depth may be associated with root activity and microbial processes (), as further supported by .
Cadmium concentrations were higher, particularly in samples from Lhze at the 0–30 cm depth. However, Cd levels tended to decrease with increasing soil depth, consistent with the findings of . The elevated Cd content in surface soil profiles of cacao plantations might primarily result from high fertilizer inputs, atmospheric deposition (mainly from rainfall), and parent material, which affect the fractions present in the topsoil (; ). In a recent study of Colombian cacao soils, an average Cd concentration of 1.43 mg·kg-1 was reported across 1,837 soil samples, well above the natural background levels typically observed in soils worldwide ().
3.2 Analysis of soil properties across profiles in Loreto localities
The biplot (Figure 2) shows a clear separation among sample groups from different localities and soil profiles (0–30 cm and 30–60 cm), indicating that soil properties vary both spatially and across profiles. The PCA accounted for 77.1% of the total variability in the data through its first two principal components (Table 2). These dimensions are sufficient to capture and visualize the majority of the data structure. Samples from Lhze at both 0–30 cm and 30–60 cm depths cluster on the right side of the plot, in the quadrant where the arrows for pH, CEC, Mg, Ca, clay, and K are also projected, indicating that these variables exert greater influence in this locality and soil profile. In contrast, samples from Arhant at 0–30 cm and 30–60 cm depth are positioned to the left, in the same direction as the arrows for acidity, sand, and aluminum (Al), suggesting higher concentrations of these variables in those soils. These results demonstrate that soil properties vary not only across localities but also along soil profiles.
FIGURE 2
TABLE 2
| Component | Sum of Squared Loadings | Percent_Variance | Cumulative_Percent |
|---|---|---|---|
| 1 | 5.74 | 57.4 | 57.4 |
| 2 | 1.97 | 19.7 | 77.1 |
| 3 | 0.92 | 9.2 | 86.3 |
| 4 | 0.68 | 6.8 | 93.2 |
PCA variance analysis in the study localities.
The Cd vector points in the opposite direction to that of pH, indicating a negative association; that is, higher pH is associated with lower Cd concentrations, whereas lower pH (more acidic conditions) corresponds to higher Cd levels. This behavior is consistent with the greater mobility of Cd in acidic soils. Likewise, the Cd vector also points in the opposite direction to that of CEC, indicating a negative correlation; soils with higher CEC tend to exhibit lower levels of available Cd, likely due to a greater adsorption capacity associated with colloidal complexes. Consequently, soils with low CEC show a stronger association with Cd (Figure 2).
The variable of interest, Cd, shows a partial correlation with CEC, Ca, and Mg, while also displaying a certain degree of independence from the dominant PCA patterns. This interpretation is supported by the factor loading values presented in Table 3. Cd has a loading of 0.2517 on PC1 and –0.039 on PC2, indicating that its variance is explained mainly by the first component and to a lesser extent by the second. Furthermore, the uniqueness value for Cd is 0.01526, the lowest among all variables analyzed. This means that 98.5% of Cd variance is well explained by the principal components, which is a highly favorable result. Table 3 further reinforces this interpretation by showing the factor loadings of each variable on the first four components. Variables such as Mg (0.395), CEC (0.385), Ca (0.385), pH (0.357), and K (0.377) exhibit high loadings on PC1, indicating that they are the main contributors to the variance explained along this axis. Meanwhile, the uniqueness values of each variable reveal the extent to which their variance lies outside the PCA model.
TABLE 3
| Variable | PC1 | PC2 | PC3 | PC4 | Uniqueness |
|---|---|---|---|---|---|
| Cd | 0.2517 | −0.039 | 0.0431 | −0.9581 | 0.01526 |
| pH | 0.3572 | 0.0856 | −0.3859 | 0.0651 | 0.71189 |
| CEC | 0.3853 | −0.0983 | −0.3003 | 0.129 | 0.73508 |
| K | 0.3772 | −0.0724 | −0.0644 | 0.0005 | 0.84835 |
| Sand | −0.3019 | 0.3305 | −0.4165 | −0.1416 | 0.6061 |
| Clay | 0.2551 | −0.3161 | 0.5862 | 0.1094 | 0.47938 |
| Ca | 0.3855 | 0.005 | −0.304 | 0.1424 | 0.73865 |
| Mg | 0.3954 | 0.053 | 0.072 | 0.0771 | 0.82973 |
| Al | −0.1254 | −0.6358 | −0.3164 | −0.0422 | 0.47812 |
| Acidity | −0.2036 | −0.6 | −0.2001 | −0.0325 | 0.55744 |
Vector analysis of the principal components (PCA) in the study localities.
Previous studies have demonstrated that soil CEC and the balance of exchangeable cations (Ca2+, Mg2+, and K+) are key factors influencing plant growth and development (Yang et al., 2024), which may be promoting cacao growth in the Lhze locality. The dynamics of the dispersive behavior of clay, in turn, determine soil characteristics such as permeability and aggregate stability, and consequently crop productivity (Zhu et al., 2020). Magnesium is an essential nutrient for cultivated plants, directly involved in photosynthesis and numerous metabolic processes; its deficiency negatively affects crop yield and quality, either directly or by impairing the utilization of other plant nutrients (). Soil pH is another critical parameter that influences a wide range of soil properties and processes, including nutrient cycling, microbial activity, inorganic carbon, and metal speciation (). It strongly affects the nutrient availability cycle and soil fertility ().
Figure 3 presents a clustered heatmap with hierarchical grouping, which enables the visualization of multivariate patterns and similarities among study localities and soil depths. This tool is useful for identifying groups of samples with similar profiles of physicochemical variables and highlighting which localities are associated with higher or lower concentrations of specific elements. Each cell in the heatmap represents a standardized value (z-score), indicating whether a sample shows a value above or below the mean for each variable. Regarding cadmium (Cd), standardized positive values are mainly observed in Lhze samples at 0–30 cm, indicating concentrations above the overall mean across localities. In contrast, samples such as Jbrllos (30–60 cm) and Arhant (0–30 cm) show negative Cd values, reflecting lower levels of this metal.
FIGURE 3
This pattern suggests that the Lhze locality, particularly in the surface layer, contains soils with higher cadmium accumulation, which is relevant for environmental and agricultural monitoring given the potential risk of bioaccumulation. Beyond cadmium behavior, the heatmap also highlights contrasting patterns in other soil variables across the evaluated localities. Clay shows elevated standardized values in Lhze samples at 30–60 cm, reflecting greater accumulation of fine particles in these soils, typically associated with higher nutrient retention capacity. Conversely, sand exhibits predominantly negative values in the same samples, confirming a finer soil texture. Exchangeable bases such as Ca and Mg tend to be more abundant in Lhze soils at 0–60 cm depths, suggesting lower leaching in this area. Similarly, K and effective cation exchange capacity show their highest values in Lhze soils at 0–60 cm, which may reflect surface layer accumulation processes or greater relative fertility at these depths.
In contrast, both soil layers from Arhant samples exhibit a clear tendency toward negative values for most variables, indicating a more degraded soil profile with lower nutrient availability. Soil pH also varies across localities, with higher values (less acidic) in Lhze, while more acidic soils are evident in Arhant and Jbrllos, a relevant pattern given its influence on the mobility of metals such as cadmium. Overall, these results demonstrate a clear differentiation among localities based on their physicochemical properties, which are further modulated by soil depth.
The higher Cd concentrations in the upper soil layers (first 30 cm sampled) may be partly attributed to the long-term accumulation of the metal in fallen leaves and husks. In cocoa farms of central Colombia, cocoa litter has been found to contain higher Cd concentrations than both cocoa beans and green leaves, with an average concentration of 85.5 mg·kg-1, indicating an intense cadmium cycling process (; ). This is further supported by studies suggesting that tree litter use as an organic amendment can indeed act as an additional source of Cd in soils (). Several soil factors influence Cd phytoavailability, including the total Cd concentration in soil, pH, organic matter content, clay mineralogy, and soil solution cations, present in the Lhze locality (; Young, 2013).
3.3 Spatial distribution of soil cadmium at two depths across study localities in the Loreto region
In the surface soil layer (0–30 cm), Cd distribution in Lhze soils exhibited significant spatial autocorrelation (Moran’s Index = 0.744; p < 0.05) and the best model’s cross-validation performance between observed and predicted values (RMSE = 0.24 and R2 = 0.50). These results indicate a moderate level of spatial predictability, enabling the identification of Cd distribution and accumulation patterns (Figures 4C1). In contrast, the Arhant (Figures 4A1) and Jbrllos (Figures 4B1) localities showed low spatial autocorrelation (Moran’s Index <0.3; p > 0.05) and limited predictive capacity (R2 < 0.30). At greater depths (30–60 cm), all three localities (Figures 4A2–C2) exhibited weak spatial structuring of Cd, with Moran’s Index values <0.2 (p > 0.05) and low coefficients of determination (R2 < 0.31). This suggests a more heterogeneous distribution, as no consistent spatial pattern of Cd distribution with respect to distance was observed.
FIGURE 4
In the surface layer (0–30 cm), the Lhze locality exhibited moderate predictive capacity (R2 ≈ 0.50), whereas Arhant and Jbrllos showed low values (R2 < 0.30). In the deeper layer (30–60 cm), all localities displayed low spatial autocorrelation and reduced predictive capacity, which can be attributed to the high small-scale variability of Cd concentrations. This variability is likely influenced by soil heterogeneity, differences in agricultural practices, and environmental processes that generate irregular distributions of the metal, thereby hindering the identification of a well-defined spatial pattern. Similar findings have been reported in cacao-producing regions of northern Peru (R2 = 0.22; ), Colombia (R2 = 0.28; ), and Costa Rica (R2 = 0.26; ), confirming that the spatial heterogeneity of Cd constrains the predictive capacity of geostatistical models in cacao systems.
Figure 4 presents the spatial distribution maps of Cd (mg·kg-1) in soils at 0–30 cm and 30–60 cm in Arhant, Lhze, and Jbrllos, generated through ordinary kriging. The maps effectively represented the spatial distribution of the metal and guided the identification of potential accumulation zones.
These findings highlight the importance of focusing attention on areas with higher Cd concentrations, as even relatively reduced cultivated areas may exceed Cd permissible limits for agricultural soils (1–3 mg·kg-1; ). In this study, only the Lhze locality surpassed the threshold of 1 mg·kg-1 (though without reaching the maximum permissible limit of 1.4 mg·kg-1), representing a potential risk in markets with stricter regulatory standards. This reinforces the need for continuous monitoring and management practices aimed at preventing Cd accumulation in the root zone and its subsequent transfer to cacao beans. The statistical evidence supports the presence of edaphic hotspots, induced by localized microenvironmental and geochemical processes such as selective adsorption in specific soil fractions or the influence of point-source contamination (Xu et al., 2025). These findings suggest that soil characteristics may directly or indirectly modulate the spatial distribution of Cd (). Overall, this study demonstrates that total cadmium (Cd) content in cacao-cultivated soils of Loreto, Peru, exhibits a highly heterogeneous spatial distribution. Concentrations exceeding 1 mg·kg-1 cannot be attributed to random variability but rather to the systematic influence of edaphic and profile-related factors (). Since metals can accumulate in agricultural soils and pose serious risks to human health, assessing soil quality is crucial to prevent potential harm associated with future contamination ().
The pathways through which cadmium is absorbed and loaded into cocoa beans have not yet been fully elucidated. Previous studies have suggested that Cd loading into cocoa beans may occur via direct uptake through the xylem rather than through phloem-mediated redistribution from leaves (). Regions of South and Central America host cocoa-growing areas, where the highest Cd concentrations in cocoa beans have been reported compared with other parts of the world, with values ranging from 0.6 to 0.8 mg kg-1 (). Likewise, reported average soil Cd concentrations of 0.39 mg kg-1in Peruvian Amazon soils; in the same sites, mean Cd concentrations in leaves ranged from 1.12 to 1.28 mg kg-1, whereas bean concentrations averaged 0.35 and 0.34 mg kg-1. Additionally, indicated that the combined effects of soil properties and cultivar resulted in high Cd concentrations in cocoa beans, exceeding the typical commercial limit of 0.6 mg Cd kg-1.
To obtain more precise and site-specific results on the translocation of Cd from soil to cocoa beans, further research more specifically focused on these processes is warranted.
3.4 Soil texture distribution in the study localities of the Loreto region
Figure 5 reveals that, contrary to the expected general homogeneity, soil textures vary markedly across localities and depths. In Lhze, clay and clay loam textures dominate both at 0–30 cm and 30–60 cm, consistent with the high values of CEC, Ca, and Mg observed in the heatmap, properties typically associated with finer-textured soils. In contrast, Jbrllos exhibits a greater proportion of coarser textures, such as sandy clay loam, in both profiles, which aligns with its lower levels of Mg, K, and CEC. Meanwhile, Arhant shows a heterogeneous mix in the 0–30 cm layer, with a notable proportion of clay and clay loam, which becomes sandier and more variable at 30–60 cm, reinforcing the heatmap results that indicated lower nutrient accumulation. This complementarity between soil texture and chemical elements underscores the importance of considering granulometry as a key structural factor directly influencing soil fertility and the dynamics of heavy metals such as cadmium.
FIGURE 5
In the cacao soils of Lhze, both at 0–30 cm and 30–60 cm, clay and clay loam textures predominate, with higher cadmium concentrations compared to the other sites. These soils also present elevated Ca and Mg contents, reflecting a positive correlation between exchangeable Ca and clay minerals (). Clay minerals are key soil components that often coexist with organic matter, forming mineral–organic associations that regulate the speciation, mobility, and bioavailability of heavy metals (Xu et al., 2022). demonstrated that the silt–clay aggregate (<0.053 mm) exhibited the highest Cd enrichment across all soils tested, consistent with the findings of . Cd preferentially binds to fine aggregates such as clay. Furthermore, fine-sized aggregates generally exhibit narrower pore diameters and higher clay content, which collectively reduce water conductivity and limit the physical mobility of Cd (). Consequently, these aggregates enhance Cd retention, explaining its consistently higher enrichment observed in the silt–clay aggregate. In addition, Cd sorption is strengthened by the presence of abundant hydrated oxides, clay minerals, and organic matter, while its mobility is further influenced by soil pH (). By contrast, the cacao soils of Jbrllos and Arhant, where sandy textures are more prevalent, showed lower Cd contents. These results are consistent with the observations of , who also reported lower Cd levels in sandy soils.
In the 0–30 cm profile (Figure 6), strong positive correlations were observed between Cd and Ca (r = 0.72), Cd and Mg (r = 0.77), and Cd and CEC (r = 0.74), suggesting that cadmium content is strongly influenced by cation exchange capacity and the presence of basic cations. An inverse relationship between Cd and sand content (r = −0.63) was also noted, consistent with the fact that sandy soils tend to retain fewer heavy metals. Among the study sites, Lhze most closely reflected these conditions, exhibiting high correlations with cations (Ca, Mg, and K), which were particularly associated with higher CEC values. In the 30–60 cm profile (Figure 7), the correlations were generally stronger, especially in Lhze, where Cd showed a strong association with K (r = 0.85) and with CEC (r = 0.53). This pattern may be attributed to greater stabilization of Cd in less mobile forms at deeper soil profiles.
FIGURE 6
FIGURE 7
In Figure 6, pH shows a positive but weak relationship with Cd across all soils; meanwhile, cation exchange capacity (CEC) emerges as the most influential factor, particularly in Jbrillos and Lhze. This indicates that Cd behavior in the 0–30 cm soil layer is more strongly dependent on soil type than on any single chemical factor. Figure 7 shows that in Arhant, Cd is clearly associated with higher CEC, even at greater depths, whereas in Jbrillos and Lhze the relationship is moderate but less consistent. Therefore, unlike the surface layer, in the 30–60 cm depth interval CEC remains more explanatory than pH, although with lower intensity. In addition, a significant and positive correlation between clay content and Ca and Mg was observed, with correlation coefficients of 0.496 and 0.685, respectively.
Calcium exhibits specific resistance mechanisms when plants are exposed to Cd stress and serves as an effective means of controlling Cd translocation from soil to rice grains (). In this study, Cd showed positive correlations with Ca, Mg, and CEC, indicating that these elements contribute to Cd immobilization in the soil by retaining this metal and preventing its transfer into cacao tissues and beans. Consequently, calcium (Ca2+) reduces Cd mobility within plants, thereby limiting its translocation from roots to shoots and decreasing overall toxicity (). Similarly, reported positive correlations between Cd with Ca, Mg, and K in cacao farms, consistent with our results. These correlations were particularly strong in Lhze, where Cd stabilization in less mobile forms was more evident at deeper soil profiles. An inverse relationship was also observed between Cd and sand content, consistent with the fact that sandy soils tend to retain fewer heavy metals. Under such conditions, Cd is highly mobilized and leached into depths between 0 and 30 cm (; ). In contrast, our study shows greater Cd stabilization in less mobile forms within deeper profiles (30–60 cm). Therefore, sand content has a significant negative correlation with Cd retention parameters ().
4 Conclusion
The Lhze locality exhibited a higher cadmium content. This result underscores the susceptibility of this locality to Cd accumulation, likely due to its combination of clayey texture, high Ca and Mg contents, high CEC, and slightly acidic pH, conditions that favor metal adsorption. Other variables also differed significantly. For instance, pH was more acidic in Arhant and closer to neutral in Lhze, which can affect Cd bioavailability, as its mobility tends to increase in more acidic soils. Similarly, CEC, Ca, and Mg values were significantly higher in Lhze, reflecting a greater cation retention capacity of the soil. Regarding the spatial distribution of Cd, in the surface profile (0–30 cm), Lhze exhibited significant spatial autocorrelation and moderate predictive capacity, whereas Arhant and Jbrllos showed low autocorrelation and limited predictability. At deeper layers (30–60 cm), all three localities displayed weak spatial autocorrelation and poor model predictability, indicating a heterogeneous distribution of Cd without a defined spatial pattern. Overall, the spatial distribution of Cd in cacao soils revealed, in most localities and depths, low autocorrelation and limited predictive capacity, attributable to high local variability. The maps represented the spatial distribution of the metal and guided the identification of potential accumulation zones. Such analyses are essential to guide agricultural management decision-making, environmental monitoring, and risk assessment of heavy metal contamination in cacao-producing areas and other crops in the Peruvian Amazon basin. These findings provide essential insights for guiding agricultural management decision-making, environmental monitoring, and metal contamination risk assessment in cacao-growing areas and other agrarian systems in the Peruvian Amazon basin. Furthermore, to promote the immobilization of cadmium from soil to plants, it is necessary to apply well-matured compost derived from cocoa residues and simultaneously establish leguminous cover crops to enhance cation exchange capacity (CEC) and soil structural stability. Through these practices, overall soil quality could be improved.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
GV-T: Formal Analysis, Methodology, Validation, Writing – original draft, Investigation. RC-R: Software, Formal Analysis, Visualization, Data curation, Writing – review and editing. NG-J: Data curation, Formal Analysis, Software, Visualization, Writing – review and editing. CM: Conceptualization, Supervision, Validation, Visualization, Writing – review and editing. JC: Formal Analysis, Funding acquisition, Investigation, Project administration, Supervision, Writing – original draft. RS: Conceptualization, Formal Analysis, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft.
Funding
The author(s) declared that financial support was received for this work and/or its publication. The research was funded by the Instituto Nacional de Innovación Agraria, within the framework of the project: Mejoramiento de los servicios de investigación y transferencia tecnológica en el manejo y recuperación de suelos agrícolas degradados y aguas para riego en la pequeña y mediana agricultura en los departamentos de Lima, Áncash, San Martín, Cajamarca, Lambayeque, Junín, Ayacucho, Arequipa, Puno y Ucayali” CUI 2487112.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
Adarme-DuranC. A.ÁgredaJ.BrandãoP. F. B.CastilloE. (2024). Cadmium availability in rhizosphere and non-rhizosphere soils in cacao farms in Santander, Colombia. Environ. Monit. Assess.196, 1–14. 10.1007/s10661-024-13301-x
2
AhmedN.ZhangB.BozdarB.ChacharS.RaiM.LiJ.et al (2023). El poder del magnesio: Descubriendo el potencial para aumentar el rendimiento, la calidad y la tolerancia al estrés de los cultivos hortícolas. Front. Plant Sci.14, 1285512. 10.3389/fpls.2023.1285512
3
AlbarrcínH. S. R.ContrerasA. E. D.HenaoM. C. (2019). Spatial regression modeling of soils with high cadmium content in a cocoa producing area of central Colombia. Geoderma Reg.16, e00214. 10.1016/j.geodrs.2019.e00214
4
ArgüelloD.DekeyrelJ.ChavezE.SmoldersE. (2022). “Gypsum application lowers cadmium uptake in cacao in soils with high cation exchange capacity only,” in A soil chemical análisis. 10.1111/ejss.13230
5
BarrazaF.MooreR. E. T.RehkämperM.SchreckE.LefeuvreG.KreissigK.et al (2019). Cadmium isotope fractionation in the soil–cacao systems of Ecuador: a pilot field study. RSC Adv.9, 34011–34022. 10.1039/c9ra05516a
6
BazánR. (2017). Manual de procedimientos de los análisis de suelos y agua con fines de riego. (1 Edn.). Available online at: https://repositorio.inia.gob.pe/items/3e246861-3056-46b4-9464-f8a175f77fe3.
7
BlommaertH.AucourA.-M.WiggenhauserM.MoensC.TeloukP.CampilloS.et al (2022). From soil to cacao bean: unravelling the pathways of cadmium translocation in a high Cd accumulating cultivar of Theobroma cacao L. Front. Plant Sci.13, 1055912. 10.3389/fpls.2022.1055912
8
BravoD.Leon-MorenoC.MartínezC. A.Varón-RamírezV. M.Araujo-CarrilloG. A.VargasR.et al (2021). The first national survey of Cadmium in cacao farm soil in Colombia. Agronomy11, 761.
9
BravoF.Arévalo-GardiniE.ArévaloL. A.BaligarV. C. (2021). Spatial distribution of cadmium in cacao (Theobroma cacao L.) plantations in Colombia. Agronomy11 (4), 761.
10
BuolS. W.SanchezP. A.CoutoW. (1982). The fertility capability soil classification system: interpretation, applicability and modification. Geoderma27 (4), 283–309. 10.1016/0016-7061(82)90019-2
11
CarrilloW.GómezE.GarcíaD. (2023). Spatial variability of cadmium in soils and its relationship with cacao bean content in Costa Rica. Environ. Monit. Assess.195 (3), 287. 10.1007/s10661-023-11817-2
12
CharryA.PereaC.RamírezK.ZambranoG.YoveraF.SantosA.et al (2025). The bittersweet economics of different cacao production systems in Colombia, Ecuador and Peru. Agric. Syst.224, 104235. 10.1016/j.agsy.2024.104235
13
ChavezE.HeZ. L.StoffellaP. J.MylavarapuR. S.LiY. C.MoyanoB.et al (2015). Concentration of cadmium in cacao beans and its relationship with soil cadmium in Southern Ecuador. Sci. Total Environ.533, 205–214. 10.1016/j.scitotenv.2015.06.106
14
ChenY.ZhangY.LiH.WeiS. (2025). Distribution characteristics of cadmium in soil aggregates and their regulating effects on Cd bioavailability. Agriculture15 (14), 1514. 10.3390/agriculture15141514
15
Corazon-GuivinM. A.Vallejos-TorresG.Vallejos-TapullimaA.Tenorio-CercadoM. Á.CaballeroW. M.MarínC.et al (2022). Rhizoglomus cacao, a new species of the glomeraceae from the rhizosphere of Theobroma cacao in Peru, with an updated identification key for all species attributed to Rhizoglomus. Nova Hedwig.115 (1-2), 99–115. 10.1127/nova_hedwigia/2022/0698
16
DaveyB. G.ConyersM. K. (1998). Determining the pH of acid soils. Soil Sci.146, 141–150. 10.1097/00010694-198809000-00001
17
DingX.ZhaoZ.XingZ.LiS.LiX.LiuY. (2021). Comparison of models for spatial distribution and prediction of cadmium in subtropical forest soils, Guangdong, China. Land10 (9), 906. 10.3390/land10090906
18
DourojeanniM. (2021). Loreto sostenible al 2021 (Lima: derecho, Ambiente y Recursos Naturales). Available online at: https://www.dar.org.pe/archivos/publicacion/Loreto2021_completo2.pdf.Primera.
19
DuH.HarataN.LiF. (2018). Responses of riverbed sediment bacteria to heavy metals: integrated evaluation based on bacterial density, activity and community structure under well-controlled sequencing batch incubation conditions. Water Res.130, 115–126. 10.1016/j.watres.2017.10.070
20
EPA (2007). Environmental protection agency, office of research and development. Environmental monitoring and support laboratory. Flame atomic absorption spectrophotometry (EPA method 7000B). Available online at: https://www.epa.gov/sites/production/files/2015-12/documents/7000b.pdf July 21, 2025).
21
GoelR.ChaturvediP.KumarS.SoniR.SuyalD. C. (2024). Editorial: hazardous pollutants in agricultural soil and environment. Front. Microbiol.15, 1411735. 10.3389/fmicb.2024.1411735
22
GramlichA.TandyS.GauggelC.LópezM.PerlaD.GonzalezV.et al (2018). Absorción de cadmio en el suelo por el cacao en Honduras. Sci. Total Environ.612, 370–378. 10.1016/j.scitotenv.2017.08.145
23
HaiderF. U.LiqunC.CoulterJ. A.CheemaS. A.WuJ.ZhangR.et al (2021). Cadmium toxicity in plants: impacts and remediation strategies. Ecotoxicol. Environmental Safety211, 111887. 10.1016/j.ecoenv.2020.111887
24
HuangF.WeiX. M.ZhuT. B.LuoZ. X.CaoJ. H. (2021). Perspectivas sobre la distribución de metales pesados disponibles en el suelo en zonas kársticas y sus factores influyentes en Guilin, suroeste de China. Forests12, 609. 10.3390/f12050609
25
KruszewskiB.WiesławO. M.KowalskaJ. (2018). Nickel, cadmium and lead levels in raw cocoa and processed chocolate mass materials from three different manufacturers. J. Food Compos Anal.66, 127–135. 10.1016/j.jfca.2017.12.012
26
KubierA.WilkinR. T.PichlerT. (2019). Cadmium in soils and groundwater: a review. Appl. Geochemistry108, 104388. 10.1016/j.apgeochem.2019.104388
27
LiZ. R.WangJ. X.AnL. Z.TanJ. B.ZhanF. D.WuJ.et al (2019). Effect of root exudates of intercropping Vicia faba and Arabis alpina on accumulation and sub-cellular distribution of lead and cadmium. Int. J. Phytoremediation21, 4–13. 10.1080/15226514.2018.1523867
28
LiQ.DengQ.FangH.YuX.FanZ.DuZ.et al (2022). Factors affecting cadmium accumulation in the soil profiles in an urban agricultural area. Sci. Total Environment807, 151027. 10.1016/j.scitotenv.2021.151027
29
LinkD. D.WalterP. J.KingstonH. M. (1998). Development and validation of the new EPA microwave-assisted leach method 3051A. Environ. Sci. Technol.32, 3628–3632. 10.1021/es980559n
30
LiuJ.FengX.QiuG.LiH.WangY.ChenX.et al (2023). Inhibition roles of calcium in cadmium uptake and translocation in rice: a review. Int. J. Mol. Sci.24 (14), 11587. 10.3390/ijms241411587
31
LoganathanP.VigneswaranS.KandasamyJ.NaiduR. (2012). Cadmium sorption and desorption in soils: a review. Crit. Rev. Environ. Sci. Technol.42, 489–533. 10.1080/10643389.2010.520234
32
LuY.GuK.WangX.ShenZ.TangC. S.ShiB.et al (2024). The role of soil structure on the cracking and cadmium leaching behavior of biochar-amended fine-grained soils. Chemosphere362, 142596. 10.1016/j.chemosphere.2024.142596
33
MaY.NingJ.YangH.ZhangL.XuC.HuangC.et al (2024). Distribution characteristics, risk assessment, and source analysis of heavy metals in farmland soil of a karst area in Southwest China. Land13 (7), 979. 10.3390/land13070979
34
MadhavS.MishraR.KumariA.SrivastavA. L.AhamadA.SinghP.et al (2024). A review on sources identification of heavy metals in soil and remediation measures by phytoremediation-induced methods. Int. J. Environ. Sci. Technol.21, 1099–1120. 10.1007/s13762-023-04950-5
35
MarkovićJ.JovićM.SmičiklasI.Šljivić-IvanovićM.OnjiaA.TrivunacK.et al (2019). Cadmium retention and distribution in contaminated soil: effects and interactions of soil properties, contamination level, aging time and in situ immobilization agents. Ecotoxicol. Environmental Safety174, 305–314. 10.1016/j.ecoenv.2019.03.001
36
MIDAGRI-Ministerio de Desarrollo Agrario y Riego.M. (2023). Perfil productivo y competitivo de los principales cultivos del sector. Consulted. Available online at: https://app.powerbi.com/view?r=eyJrIjoiYjYwYTk5MDgtM2M0MS00NDMyLTgzNDEtMjNhNjEzYWQyOTNlIiwidCI6IjdmMDg0NjI3LTdmNDAtNDg3OS04OTE3LTk0Yjg2ZmQzNWYzZiJ9 (Accessed November 08, 2025).
37
MINAM-Ministerio del Ambiente (2015). Zonificación ecológica y económica de la provincia de Alto Amazonas, departamento de Loreto (Vol. I: Diagnóstico físico-natural). Dirección General Ordenamiento Territ. Ambient. – Minist. del AmbienteAvailable online at: https://sinia.minam.gob.pe/sites/default/files/siar-loreto/archivos/public/docs/1633.pdf (Accessed July 16, 2025).
38
MINAM. Ministerio del Ambiente (2017). Decreto Supremo N° 011-2017-MINAM Estándares de Calidad Ambiental (ECA) Para Suelo. Available online at: https://sinia.minam.gob.pe/normas/aprueban-estandares-calidad-ambiental-eca-suelo-0 (Accessed July 11, 2025).
39
Ministerio de Agricultura y Riego (MINAGRI) (2019). “Dirección General de Políticas Agrarias/Dirección de Estudios Económicos e Información Agraria,” in Requerimientos Agroclimáticos del cultivo de cacao (Accessed July 22, 2025).
40
Ministerio del Ambiente. MINAM (2014). Guía para el muestreo de suelos. Available online at: https://www.minam.gob.pe/calidadambiental/wpcontent/uploads/sites/22/2013/10/GUIA-PARA-EL-MUESTREO-DE-SUELOSfinal.pdf. (Accessed July 22, 2025).
41
MosleyL. M.RengasamyP.FitzpatrickR. (2024). Soil pH: techniques, challenges and insights from a global dataset. Eur. J. Soil Sci.75 (6), e70021. 10.1111/ejss.70021
42
NelinoR. (2021). Cadmium in soil and cacao beans of Peruvian and South American origin. Rev. Fac. Nac. Agron. Medellín74 (2), 9499–9515. 10.15446/rfnam.v74n2.91107
43
NOM-021-RECNAT-2000 (2002). Natural resources and environment secretary. Mexican official standard. Specifications of fertility, salinity, and soil classification. Study, Sampl. Analysis. Available online at: https://www.ordenjuridico.gob.mx/Documentos/Federal/wo69255.pdf (Accessed October 20, 2025).
44
OladimejiT. E.OyedemiM.EmetereM. E.AgboolaO.AdeoyeJ. B.OdunlamiO. A. (2024). Review on the impact of heavy metals from industrial wastewater effluent and removal technologies. Heliyon10 (23), e40370. 10.1016/j.heliyon.2024.e40370
45
OlivaM.RubioK.EpquinM.MarloG.LeivaS. (2020). Cadmium uptake in native cacao trees in agricultural lands of Bagua, Peru. Agronomy10 (10), 1551. 10.3390/agronomy10101551
46
PanS. F.JiX. H.XieY. H.LiuS. H.TianF. X.LiuX. L. (2022). Influence of soil properties on cadmium accumulation in vegetables: thresholds, prediction and pathway models based on big data. Environ. Pollut.304, 119225. 10.1016/j.envpol.2022.119225
47
Pulgar-VásquezS.Arévalo-GardiniE.BravoF.CantoM.OrtizC. (2023). Cadmium concentration mapping in cocoa-growing areas of Peru using machine learning approaches. Environ. Sci. Pollut. Res.30, 68235–68249. 10.1007/s11356-023-27687-w
48
R Core Team (2024). R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing. Available online at: https://www.R-project.org/.
49
RobinsonN. J.BenkeK. K.NorngS.KitchingM.CrawfordD. M. (2017). Improving the information content in soil pH maps: a case study. Eur. J. Soil Sci.68 (5), 592–604. 10.1111/ejss.12452
50
Rojas-BriceñoN. B.GarcíaL.Cotrina-SánchezA.GoñasM.SalasR.SilvaJ. O.et al (2022). Land suitability for cocoa cultivation in Peru: AHP and MaxEnt modeling in a GIS environment. Agronomy12 (12), 2930. 10.3390/agronomy12122930
51
ScaccabarozziD.CastilloL.AromatisiA.MilneL.BúllonA.Muñoz-RojasM. (2020). Soil, site and management factors affecting cadmium concentrations in cacao-growing soils. Agronomy10 (6), 806. 10.3390/agronomy10060806
52
ShabtaiI. A.WilhelmR. C.SchweizerS. A.HöschenC.BuckleyD. H.LehmannJ. (2023). Calcium promotes persistent soil organic matter by altering microbial transformation of plant litter. Nat. Commun.14, 6609. 10.1038/s41467-023-42291-6
53
ShenQ.XiangJ. M.ZhangM. K. (2022). Distribution and chemical speciation of heavy metals in various size fractions of aggregates from zonal soils. Int. J. Environ. Anal. Chem.102, 4272–4287. 10.1080/03067319.2020.1781840
54
SINIA-Sistema Nacional de Informacion Ambiental (2021). Diagnóstico Ambiental Regional de Loreto. Sist. Nac. Inf. Ambient. – Minist. del Ambiente. Available online at: https://sinia.minam.gob.pe (Accessed July 29, 2025).
55
SolórzanoR.LlerenaR.MejíaS.CruzJ.QuispeK. (2025). Spatial distribution of cadmium in avocado-cultivated soils of Peru: influence of parent material, exchangeable cations, and trace elements. Agriculture15 (13), 1413. 10.3390/agriculture15131413
56
United States Environmental Protection Agency (1994). Method 200.7: determination of metals and trace elements in water and wastes by inductively coupled plasma-atomic emission spectrometry. Available online at: https://www.epa.gov/esam/method-2007-determination-metals-and-trace-elements-water-and-wastes-inductively-coupled (Accessed January 26, 2026).
57
United States Environmental Protection Agency (1996). Method 3050B: acid digestion of sediments, sludges, and soilsAvailable online at: https://www.epa.gov/sites/production/files/2015-06/documents/epa-3050b.pdf (Accessed January 26, 2026).
58
Vallejos-TorresG.TorresS. C.Gaona-JimenezN.SaavedraJ.TuestaJ. C.TuestaO. A.et al (2022). The combined effect of arbuscular Mycorrhizal Fungi and compost improves growth and soil parameters and decreases cadmium absorption in cacao (Theobroma cacao L.) plants. J. Soil Sci. Plant Nutr.22, 5174–5182. 10.1007/s42729-022-00992-9
59
Vallejos-TorresG.Gaona-JimenezN.AlvaA.ParedesC.LozanoA.Saavedra-RamírezJ.et al (2023). Cadmium uptake and mycorrhization by cacao clones in agroforestry and monoculture systems of Peruvian Amazon. Bioagro35 (3), 237–246. 10.51372/bioagro353.7
60
VanderschuerenR.ArgüelloD.BlommaertH.MontalvoD.BarrazaF.MauriceL.et al (2021). Mitigating the level of cadmium in cacao products: reviewing the transfer of cadmium from soil to chocolate bar. Sci. Total Environ.781, 146779. 10.1016/j.scitotenv.2021.146779
61
VanderschuerenR.WantiezL.BlommaertH.FloresJ.ChavezE.SmoldersE. (2023). Revealing the pathways of cadmium uptake and translocation in cacao trees (Theobroma cacao L.): a 108Cd pulse-chase experiment. Sci. Total Environment869, 161816. 10.1016/j.scitotenv.2023.161816
62
Villalaz-PérezJ. A.CasanovesF.Villarreal-NúñezJ. E.Santo-PinedaA.Gutiérrez-LezcanoA.MerinoA. (2024). Cadmium concentration in cocoa beans produced in agroforestry systems of small producers in Panama. Span. J. Agric. Res.22 (1), e0301. 10.5424/sjar/2024221-20405
63
VillegasC. L.QuirogaN. J.DávalosA.PizanoC. (2024). Reducing cadmium in cocoa (Theobroma cacao) cultivars: analysis of advantages and limitations of different strategies and their potential application in Colombia. Int. J. Agron.2024 (1), 5342167. 10.1155/ioa/5342167
64
WadeJ.Ac-PanganM.FavorettoV. R.TaylorA. J.EngesethN.MargenotA. J. (2022). Drivers of cadmium accumulation in Theobroma cacao L. beans: a quantitative synthesis of soil-plant relationships across the Cacao belt. PLoS ONE17 (2), e0261989. 10.1371/journal.pone.0261989
65
WangQ. Y.WuM. X.HuN. W.DengB. L.WangT. Y.YangX. T.et al (2024). Tracing the vertical migration of exogenous cadmium in soil by seasonal freeze-thaw event using rare Earth elements. Sci. Total Environ.946, 174147. 10.1016/j.scitotenv.2024.174147
66
WangJ.YangJ.ZhaoC.TianX.ZhaoX.ZhaoW.et al (2025). Revealing influencing mechanisms and spatial pattern of soil cadmium through geodetector and spatial analysis. Land14 (10), 1975. 10.3390/land14101975
67
WdowczykA.TomczykP.Szymańska‐PulikowskaA.WiatkowskaB.Rosik‐DulewskaC. (2023). Copper and cadmium content in Polish soil: analysis of 25‐year monitoring study. Land Degrad. Dev.34 (16), 4857–4868. 10.1002/ldr.4814
68
WickhamH. (2016). ggplot2: elegant graphics for data analysis. New York: Springer-Verlag. 978-3-319-24277-4. Available online at: https://ggplot2.tidyverse.org (Accessed July 10, 2025).
69
XuW.LiC.ZhuJ. M.BuH.TongH.ChenM.et al (2022). Adsorption of cadmium on clay-organic associations in different pH solutions: the effect of amphoteric organic matter. Ecotoxicol. Environ. Saf.236, 113509. 10.1016/j.ecoenv.2022.113509
70
XuH.WangH.CrootP.LiuJ.LiY.BeiyuanJ.et al (2025). Investigation of spatially varying relationships between cadmium accumulation and potential controlling factors in the topsoil of Island of Ireland based on spatial machine learning approaches. Environ. Res.275, 121466. 10.1016/j.envres.2025.121466
71
YangM.ZhouD.HangH.ChenS.LiuH.SuJ.et al (2024). Effects of balancing exchangeable cations Ca, Mg, and K on the growth of tomato seedlings (Solanum lycopersicum L.) based on increased soil cation exchange capacity. Agronomy14 (3), 629. 10.3390/agronomy14030629
72
YoungS. D. (2013). “Chemistry of heavy metals and metalloids in soils,” in Heavy metals in soils. Environmental pollution. Editor AllowayB. (Dordrecht: Springer), 22, 51–95. 10.1007/978-94-007-4470-7_3Environ. Pollut.
73
YuH.ZhaT.ZhangX.MaL. (2019). Vertical distribution and influencing factors of soil organic carbon in the loess Plateau, China. Sci. Total Environ.693, 133632. 10.1016/j.scitotenv.2019.133632
74
ZhangX.YangM.YangH.PianR.WangJ.WuA. M. (2024). The uptake, transfer, and detoxification of cadmium in plants and its exogenous effects. Cells13 (11), 907. 10.3390/cells13110907
75
ZhuY.MarchukA.BennettJ. M. (2020). Ionicity of clay–cation bonds in relation to dispersive behavior of Mg and K soil clays as influenced by pH. Clays Clay Minerals68 (6), 588–600. 10.1007/s42860-020-00100-x
76
ZhuangX.WanH.WangH.QinS.HeJ.LyuD. (2023). Characteristics of cadmium accumulation and tolerance in apple plants grown in different soils. Front. Plant Sci.14, 1188241. 10.3389/fpls.2023.1188241
Summary
Keywords
agricultural management, edaphic factors, heavy metals, physicochemical properties, soil contamination
Citation
Vallejos-Torres G, Chuchon-Remon R, Gaona-Jimenez N, Marín C, Cruz J and Solórzano R (2026) Geospatial distribution of cadmium in soil profiles of cacao (Theobroma cacao L.) plantations in the Peruvian Amazon basin. Front. Environ. Sci. 14:1710127. doi: 10.3389/fenvs.2026.1710127
Received
21 September 2025
Revised
03 February 2026
Accepted
05 February 2026
Published
25 February 2026
Volume
14 - 2026
Edited by
Oliver Matthias Wiche, Zittau/Görlitz University of Applied Sciences, Germany
Reviewed by
Juan Manuel Trujillo-González, University of the Llanos, Colombia
Precious Okoroafor, City Department of Climate Protection and Adaptation, Germany
Viktoriia Lovynska, Dnipropetrovsk State Agrarian University, Ukraine
Updates
Copyright
© 2026 Vallejos-Torres, Chuchon-Remon, Gaona-Jimenez, Marín, Cruz and Solórzano.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Geomar Vallejos-Torres, gvallejos@unsm.edu.pe
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.