Soils are essential for agricultural purposes as they are the medium through which crops are grown. Soil quality is a crucial factor for human health, animal husbandry and the sustainability of all ecosystems. Globally speaking, soil quality is affected by the presence of potential toxic elements (PTEs), which is largely due to anthropogenic activity, particularly by agriculture use ^{[1], [2], [3]}. Indeed, soil may become polluted by heavy metals (HMs) and metalloids accumulating after receiving a significant number of pollutants from different land applications of fertilizers, animal manure, sewage sludge, pesticides and wastewater irrigation. Many farmers add large quantities of fertilizers to soils in intensive farming systems to provide adequate N, P and K supplies for crop growth. These fertilizers normally contain trace amounts of heavy metals (e.g., Cd, Pb, rare earths, etc.) that ultimately increase in soil ^[4]. For example, land application of biosolid materials is a common practice in many countries, normally by urban populations ^[5]. Thus, several common pesticides and fertilizers are fairly extensively employed in agriculture and horticulture ^[6], while certain animal waste types, such as poultry, cattle and pig manures, are also used in this way ^[7]. All this poses serious growing concern about the potential risk to human health via direct intake, bioaccumulation through the food chain and impacts on the ecological system.

Three different agricultural fertilizer categories are normally established: macronutrient fertilizer to supply major nutrients (N, P, K); macronutrient fertilizer to supply secondary nutrients [calcium (Ca), magnesium (Mg) and sulfur (S)] and micronutrients [copper (Cu), zinc (Zn), nickel (Ni), manganese (Mn), iron (Fe), molybdenum (Mo), boron (B), chlorine (Cl), sodium (Na), cobalt (Co) and silicon (Si)]. There is another classification based on essentiality of nutrients, where metals may be classified in this way: B, Cu, Fe, Mn, Mo and Zn as essentials; Co, Ni and V as beneficial; Cd, Cr, Hg and Pb as toxic. So, it should be noted that essential and beneficial elements can become toxic at high concentrations ^[8]. Heavy metals can be classified into two categories: essential (biometals) and nonessential (toxic metals). Essential metals (Co, Cr, Cu, Fe, Mn, V, Zn) at high levels can also have toxic effects on organisms, while toxic metals (Cd, Pb, Al) are toxic to human health and the environment even at low concentrations.

To understand this behavior, we must bear in mind that heavy metals in soils can appear in several forms: interchangeable (bound to organic and inorganic components), as soluble ionic forms in structural components of soil particles and in insoluble and/or precipitated states. Only the first form can be generally absorbed and used by plants in such a way that they also constitute an important point of connection between these elements and humans through the food chain and represent a human health risk ^[9]. The environmental risk of heavy metals depends on several factors, such as their total contents and soil characteristics. In any case several authors have shown that agricultural activity has become the main source of heavy metals in soils ^{[3], [10]-[13]}.

According to Mahecha-Pulido et al. ^[14], soil studies in Colombia related to heavy metals have been conducted mainly in the Andean region. They have focused on the Departments of Cundinamarca, Antioquia, Valle del Cauca, Boyacá, Santander, Cauca and Nariño, and also on the Caribbean region. Here analyses have been performed to focus on the toxicological risk due to heavy metal concentrations in the middle and lower valleys of the Sinú River in the Department of Córdoba.

In the Orinoquia region, which is considered the main agricultural pantry of Colombia, very few studies have established the concentrations of these elements, or monitoring and follow-up plans in pursuit of sustainable soil resource use ^{[3], [15]-[17]}. Since this region represent unique soil and climate conditions on a global scale, our main objective was to take a look at the possible role of three agricultural production systems (oil palm, pastures, semi-annual crops) in the accumulation of some HMs in acid soils at Piedemonte Llanero in Colombia. In this way, the primary goals of this study were to: (1) provide a database for the studied HMs; (2) evaluate the effect of the agricultural production system on the concentration of metals; (3) gain a better understanding of the ecological risk assessment for the HMs that will contribute to the sustainable management of this vital hyperhumid area. For this purpose, the heavy metal concentrations in the soils of these three agricultural production systems were first determined, and later evaluated by means of the geo-accumulation index (I-geo), the contamination factor (CF) and the pollution load index (PLI). The ultimate purpose was for this information to become a basic tool to establish monitoring and follow-up processes in pursuit of sustainable soil resources management.

As a consequence of the developing under tropical conditions, most of these soils have limitations for agriculture, such as acid, deep and highly weathered soils with low nutrient availability and high exchangeable acidity (Al³⁺), because they result in low cation exchange capacity and poor P availability ^[30].

Indeed, the granulometric distribution results (Table 1) showed that the soils cultivated with pastures and oil palm have a loamy-clayey texture, while the soils with semi-annual crops are loamy-sandy, loamy-clayey and loamy-clayey-sandy. These results agree with those reported by Trujillo-González et al. ^[31] and IGAC ^[32]. The soils with pastures presented a loam-clay-sandy (F-Ar-A) texture, the % of sand was 56.15, with 21.15 % silt and 22.70 % clay. In the oil palm crops, the results were 42.14 % sand, 26.64 % silt and 31.22 % clay for a loam-clay texture (F-Ar). In the semi-annual crops, a loam texture was found with 43.86 % sand, 30.33 % silt and 25.76 % clay.

Table 1

Source: created by the authors.

The soil pH for all the samples was strongly acidic (Table 1), with a mean of 4.8. Soils below pH 6.5 are considered acidic. According to the Kruskal-Wallis H test, there were no significant differences (p<0.05) between the pH values and the soils with the pasture, oil palm and semi-annual crop production systems. Mahecha-Pulido et al. ^[11] and Trujillo-González et al. ^[31] obtained similar soil pH values in the Piedemonte Llanero region. Organic matter presented variability with a CV % of 39.7 % and a mean of 1.73 (Table 1). For this variable, no significant difference (p <0.05) appeared in any of the studied soils.

In light of all this, carrying out fertilization and soil amendment practices seems necessary to make soil suitable for major crop development, to provide plants with the necessary nutrients and to lower soil acidity. Unquestionably fertilizer overuse can also provide soil pollution because it increases the concentration of a particular element or substance of environmental interest above its naturally-occurring concentration in soil.

The studied HMs have the same sequence in the pasture, oil palm and semi-annual crop systems and come in the following descending sequence: Mn > Zn > Cr > Ni > Cu > Pb > Cd. The Kruskal-Wallis H test showed no significant difference at 0.05 among the studied soils depending on the three crop types. A comparative analysis was also carried out using a box-plot for the three production systems (pastures, oil palm, semi-annual crops), which presented the highest concentrations of HMs, followed by the oil palm-cultivated soils (Figure 2). These findings can be explained by the continuous phosphorous and nitrogenous fertilizer applications in these production systems.

Figure 2 Source: created by the authors

Heavy metal pollution showing low CV is associated with natural sources, and high CV is typically sourced from human activities. Based on this criterion and the calculated CVs of the analyzed species, the CVs of Cr, Cu, Ni and Pb in the farmland soils in the study area were almost all between 25 % and 50 %, indicating moderate variability for these elements.

Mn, Zn and Cd showed a high variability for some crops, concerning their spatial distributions; this indicates that the concentrations of Mn, Zn and Cu vary significantly from one sampling site to the other, what we attribute to be affected by extrinsic factors, such as agricultural activities. In all cases, CV % exceeded 30 %, which indicates the wide variability of concentrations in the soils of each production system. García-Martínez and Poleto ^[24] and Trujillo-González and Torres-Mora ^[33] mention that this is due to the high heterogeneity of the distribution of metals in soil, which responds to these elements’ typical behavior.

The Zn in these agricultural soils obtained a mean value of 27.62 (mg kg^-1) for the pasture production system, 32.79 (mg kg^-1) for oil palm and 38.23 (mg kg^-1) for semi-annual crops. These values are lower than those reported by Mahecha-Pulido et al. ^[11] for the same region with 58.6 (mg kg^-1) but come close to the reference value of 28.2 (mg kg^-1) obtained by Trujillo-González et al. ^[28]. The Cr concentrations in these soils were 25.68 (mg kg^-1) for pastures, 25.15 (mg kg^-1) for oil palm and 26.91 (mg kg^-1) for semi-annual crops. These values are higher than those reported by Mahecha-Pulido et al. ^[11] of 11.9 (mg kg^-1) but come close to the reference value of 21.1 (mg kg^-1) ^[28] for the same region. The Cu values herein obtained were 10.16 (mg kg^-1) for the pasture system, 11.42 (mg kg^-1) for oil palm and 13.06 (mg kg^-1) for semi-annual crops, which comes close to the reference value of 9.9 (mg kg^-1) for the Piedemonte Llanero ^[28].

The mean values of Ni were 10.37 (mg kg^-1) for pastures, 11.82 (mg kg^-1) for oil palm and 13.80 (mg kg^-1) for semi-annual crops. The soils with pastures came close to the reference of 10.2 (mg kg^-1) ^[28]. With Pb, the three production systems remained below the reference value of 11.3 (mg kg^-1) ^[28], with 9.24 (mg kg^-1) for pastures, 9.46 (mg kg^-1) for oil palm and 9.69 (mg kg^-1) for semi-annual crops. Finally, the Cd concentrations in the pasture system were 0.21 (mg kg^-1), 0.19 (mg kg^-1) in oil palm and 0.26 (mg kg^-1) in semi-annual crops. These values are lower than those reported by Yacomelo ^[33] in Candelaria (Atlantic) with a concentration of 1.29 (mg kg^-1), and one of 1.66 (mg kg^-1) in Manatí (Atlantic) and the reference value of 0.3 (mg kg^-1) for the Piedemonte Llanero ^[28].

This study, conducted to provide baseline information on the spatial variability and a risk assessment of selected heavy metals on the farmlands of the Piedemonte Llanero in Colombia, represent a preliminary advance in its knowledge. In this study, selected HMs (Cu, Cd, Cr, Mn, Ni, Pb, Mn and Zn) in surface soils of on farmlands at Piedemonte Llanero in Colombia were investigated by analytical methods; Also, their potential ecological risks were evaluated. The average HM concentrations in the collected topsoil samples were overall: Mn (110.5 mg kg^-1), Zn (31.93 mg kg^-1), Cr (8.85 mg kg^-1), Ni (11.68 mg kg^-1), Cu (11.28 mg kg^-1), Pb (9.42 mg kg^-1); Cd (0.21 mg kg^-1). The study represents a preliminary advance in perhumid climate and acid soils. Despite concentrations generally being below or coming close to the reference value proposed for the Piedemonte Llanero region, the application of the I-geo, the CF, the PLI factor and NIPI index enabled us to indicate that these HMs do not currently pose environmental. Therefore, it is evident that the soil samples at the studied sites were safe for agricultural and environmental purposes. The results of this study are expected to shed light on the community’s understanding, and to enable the City Council to monitor environmental quality and to take appropriate actions at Piedemonte Llanero in Colombia.