Cadmium in cacao: mitigation strategies to reduce Cd uptake.

Since 2019, the European Union (EU) has implemented a new regulation that limits the amount of cadmium (Cd) in chocolates and in derived products to protect consumer health. Cadmium concentrations in cacao bean are notably higher in cacao grown in Latin America compared to cacao grown elsewhere, likely because the young Andean soils are naturally more enriched in Cd. About 25% of Ecuadorian cacao is sold to Europe. Thus, this regulation could affect the Ecuadorian cacao export industry, which is worth more than 800 million dollars per year. Approximately 70% of the cacao production areas in Ecuador are in the hands of small-scale low-income farmers, hence the EU regulation could also affect the livelihood of Ecuadorian cacao farmers negatively. The objective of this dissertation was to identify the spatial distribution of Cd concentrations in Ecuadorian cacao, to identify soil or agronomic factors that affect Cd uptake in cacao and to find a mitigation strategy based on agricultural practices to reduce Cd concentration in cacao beans during the crop production. During the first stage of this work, a nationwide survey was conducted. A total of 560 paired samples (soil, leaves and beans) were collected from 159 cacao farms distributed across the cacao producing areas of Ecuador. Soils and plant tissues were analyzed, and information regarding agronomic practices was gathered through questionaries. A bean Cd concentration threshold of 0.6 mg Cd kg-1 was selected that allows to meet the EU limit in cacao products. That threshold was exceeded in 45% of the samples, and the nationwide mean Cd concentration was 0.90 mg Cd kg-1. Multiple regression analyses showed that bean Cd increased with increasing total soil Cd concentration and decreasing pH and organic matter content in soils (SOM), the latter two factors affecting soil Cd bioavailability. This suggested that modifying these soil factors could reduce Cd uptake in cacao. Bean Cd concentration decreased a factor of 1.4 as the age of the orchard increased from 4-40 years. Bean Cd concentration was inconsistently affected by genotype (CCN-51 vs Nacional), pruning or application of fertilizers. Kriging interpolation allowed to construct the first map of Cd in Ecuadorian cacao illustrating "hotspots" where mean Cd concentration in cacao beans exceed the threshold; these areas overlap those where soils are naturally enriched with Cd with very few cases of soil contamination. There is a large spatial variability in Cd concentrations, even among trees within the same field. However, a second independent database with 640 georeferenced data collated by the Ecuadorian government was mapped and showed that regions with hotspots and regions where cacao is below the limit largely agree with the map of this study, confirming that the map of this study is a robust indicator for regional mean Cd concentration in beans. The second stage focused on understanding how amending lime at the surface of soils can affect Cd uptake in cacao plants. Lime increases soil pH and lowers soil Cd bioavailability, but lime cannot be ploughed in soil on established cacao orchards. Surface applied lime does not penetrate the soil and the net effect of liming needs process understanding. A pot experiment was conducted where pots were filled with topsoil only or topsoil over subsoil. Additionally, pots were fully limed or partially limed (top layer only). Finally, the soil was spiked with 108Cd in several combinations to trace the provenance of Cd. About 70% of root biomass was present in the top layer, and this was not changed by any of the lime or soil provenance treatments. A complete liming of the soil reduced Cd concentrations by factor 1.7, corroborating existing knowledge that lime can successfully reduce Cd uptake. In contrast, liming the top layer reduced leaf Cd concentration by a factor of 1.2 only, a marginal effect suggesting weak effects of lime in the field. The isotope data showed that liming the top layer increased the Cd uptake from the bottom layer compared to the unlimed control. Liming decreases not only Cd availability but also that of nutrient analogues Zn and Mn, hence lower availability of the analogues may induce compensating effects on subsurface roots. This suggests that lime application in combination with micronutrients addition may be a suitable strategy. The third phase of this work focused on the effect of gypsum (CaSO4) application to soil on Cd uptake. Gypsum is often used in acid soils for subsurface liming; it is soluble and penetrates in soil. However, it is not a lime product but a product that enhances subsurface Ca2+ concentrations that may counteract effects of toxic Al3+ on soil and, potentially, on Cd2+ uptake. First, a pot experiment was conducted using six Ecuadorian soils with contrasting properties, including soil solution Ca2+ concentrations. The soil was amended or not with gypsum at a dose equivalent to 4 Mg ha-1 and grown with cacao seedlings for 115 days. Gypsum application decreased leaf Cd concentration in only one soil with none to even small enhancing effects in other soils. The factor by which Cd concentrations in the leaf are reduced by gypsum markedly increased with increasing cation exchange capacity (CEC) as well as with native Ca in soil solution, i.e., gypsum is only effective in high CEC or high Ca soils. Soil solution analysis suggests that in sandy soils, where there is low buffer capacity (or low CEC) and low Ca in solution, Cd2+ is more readily mobilized when gypsum dissolves into the soil solution and, hence, the Cd2+/Ca2+ ion activity ratio in soil solution remains similar with gypsum application with minor effects on Cd uptake. In contrast, in high CEC soils, the soil is saturated with Ca2+, and the additional Ca2+ has smaller effects on mobilizing Cd2+, yielding a decreasing Cd2+/Ca2+ ratio in soil solution with gypsum application and, hence, decreasing Cd uptake. Two field trials had been established at the start of this work to test the effectiveness of soil amendments, and these were monitored for 30 months. The field trials were located in two Ecuadorian farms with contrasting soil properties, i.e., pH (acid and pH neutral) and SOM. The treatments tested consisted of lime and gypsum at 2 and 4 Mg ha-1 y-1and compost at 25 and 50 Mg ha-1. Soil, leaf, and bean samples were collected regularly. At the end, soil cores were collected down to 100 cm. Lime significantly increased topsoil pH, but this was undetectable at depths >20 cm after 22 months. None of the lime or gypsum applications reduced leaf Cd concentrations at soil pH with neutral conditions. On the acid soil, leaf Cd concentrations decreased by lime application, and the reduction factor gradually rose to 1.4 after 30 months suggesting that internal Cd stocks in the tree or in the rooted zone explain the delayed response. Compost applications lowered leaf Cd concentration by a factor of 1.2 (p≤0.05) after 22 months, but this effect disappeared at 30 months. Bean Cd concentrations were unaffected by any of the treatments at 22 months after application (acid soil) or 30 months (pH neutral soil), Leaf Cd concentrations were significantly affected by the time of sampling and lower leaf Cd concentrations were found at the time of flowering/pod filling. Hence, the effects of soil amendments might be masked by time or plant phenological stages that affect the internal mobilization of mineral elements such as Cd. A laboratory study showed that the combination of lime with compost largely enhances lime penetration in soil compared to lime only, suggesting that a mixed treatment could be advocated as a soil amendment. In conclusion, this study identified the cacao producing areas in Ecuador where bean Cd concentrations exceed the selected threshold of 0.6 mg kg-1. We confirmed that the most important soil factors affecting Cd uptake in cacao are pH and soil organic carbon with minor effects of agronomic practices. Soil amendments reducing soil Cd bioavailability, even with a modest factor of 1.4, could have far-reaching consequences for the livelihood of Ecuadorian cacao farmers. However, the application of soil amendments to lower Cd uptake in the field faces severe limitations. These are the challenges to change subsoil properties and reach deeper roots and the large variability of soil Cd bioavailability among trees within the field, which reduces the statistical power to detect effects of, for example, factor 1.5 reduction in the field. None of the field trials yet identified suitable management techniques to lower bean Cd concentrations, except for lime in acid soils which at least affected leaf Cd concentrations within 2.5 years after application. The weight of evidence from experimental studies however suggests that applied combinations of lime with compost or lime with micronutrients such as Zn or Mn may be candidates for alternative soil treatments which should be tested in the future.

Publication year:2021

Accessibility:Open