Journal Article

Mechanistic multi-surface models predict soil geochemical drivers of cadmium accumulation in cacao

author(s)

Wiersma, Wietse

themes
  • Multifunctional landscapes
regions
  • Colombia
  • Ecuador
original file download
full citation

Wiersma, W.; Dekeyrel, J.; Chavez, E.; Comans, R.N.; Smolders, E.; Groenenberg, J.E.; Pulleman, M.M. (2026) Mechanistic multi-surface models predict soil geochemical drivers of cadmium accumulation in cacao. Geoderma 468: 117786. ISSN: 0016-7061

• Measured and predicted solution Cd concentrations are able to predict bean Cd.
• The soil Cd buffering capacity is positively related to the Cd soil–plant transfer factor.
• Simplification of MSMs did not affect their predictive power for bean Cd content.
• The MSMs emerged as powerful tools to understand Cd transfer in soil-cacao systems. Geochemical multi-surface models (MSMs) remain underexplored as tools to predict and provide insights into the uptake of heavy metals by plants.
Our study is situated in the context of food safety regulations on cadmium (Cd) in cacao products. Based on a set of 188 paired soil–plant samples representing diverse cacao growing conditions in Ecuador and Colombia, we quantified the ability of MSMs with various levels of model complexity to predict Cd accumulation in cacao beans and leaves. We related the transfer factor (ratio of Cd bean or Cd leaf to Cd soil ) to soil geochemical mechanisms that control the solid-solution partitioning and speciation of Cd. We found that most variation in bean Cd (up to 73%) and leaf Cd (up to 82%) could be explained by indicators of directly available Cd in the soil solution, either measured in a 0.001 M Ca(NO 3 ) 2 extraction or predicted with the MSM. Moreover, plant Cd content was more strongly related to Cd that was electrostatically (weakly) bound by clay and organic matter, than to Cd specifically bound to organic matter or metal-oxides, which we attribute to the capacity of soils to buffer the Cd concentration in the soil solution. The pH was identified as the key factor governing solid-solution partitioning, speciation and soil–plant Cd transfer. Informed MSM simplification steps had a negligible influence on model performance or the predictive ability for Cd in cacao. We conclude that employing MSMs is a feasible and insightful approach for environmental risk assessment and for targeting Cd mitigation strategies through soil management practices.