Mercury biomagnification emerges across a Brazilian savanna food web

Mercury concentrations increased across the surface-dwelling food web of a Brazilian Cerrado reserve, with predatory and other high-trophic-position invertebrates carrying substantially more of the metal than litter, soil and plant-feeding organisms. In the study at Águas Emendadas Ecological Station, total mercury in leaf litter averaged 51.7 ± 5.6 nanograms per gram, compared with 34.9 ± 13.5 nanograms per gram in soil. Individual invertebrate samples ranged from 0.095 to 970.70 nanograms per gram, reaching concentrations about 15 times higher than those measured in the two environmental compartments. The results indicate pronounced bioaccumulation and a pattern consistent with biomagnification, although the authors caution that feeding habit alone does not determine how much mercury an organism accumulates.

The research combined environmental sampling with taxonomic identification and statistical analysis of surface-active soil invertebrates collected in eight gallery-forest sites within the 10,500-hectare ecological station in the Federal District. Each 100-square-metre sampling unit contained 10 pitfall traps, whose combined catches formed one composite sample. Traps were deployed from May through July 2021 and collected weekly during seven campaigns. The organisms were identified, counted, measured, freeze-dried and homogenised, while soil and litter were collected from the centre of each sampling unit. Total mercury was measured by atomic absorption spectrometry, with certified reference materials used to verify analytical accuracy and recovery rates of 81% to 103%.

Mercury was significantly more concentrated in litter than in soil, a difference associated with paired statistical testing that yielded t = 4.402, seven degrees of freedom and p = 0.0032. The authors regarded the forest canopy as a plausible interception surface for atmospheric mercury, allowing mercury captured by vegetation to enter the soil compartment through litterfall. Sampling took place in gallery forests, a Cerrado formation that follows watercourses. Their dense canopies limit direct deposition to the ground, while prolonged leaf exposure, reduced litter leaching and shelter from sunlight can help retain deposited mercury. These mechanisms are consistent with the observed litter enrichment, but the comparison does not by itself identify the precise source or chemical pathway responsible.

Soil organic matter was a significant predictor of mercury concentration, accounting for 45.5% of the observed variation in the regression analysis, with p = 0.039. The estimated slope indicated that soil mercury increased by 0.537 units for each unit increase in organic matter content. Soil pH showed no significant relationship with total mercury. The authors connected the organic-matter association with mercury’s affinity for sulfur-bearing functional groups, including thiol and disulfide sites, which can promote binding of inorganic mercury in soil. They also noted that litter decomposition can transfer captured mercury into the underlying soil. Because only eight paired sampling units supplied the environmental comparison, the regression and soil–litter contrast should be interpreted within the scope of this small, site-specific field survey.

The invertebrate collections spanned 15 orders and included ants, termites, spiders, scorpions, beetles, flies, wasps, crickets and other epigeic fauna. Among orders with enough observations for comparison, termites had the highest mean mercury concentration at 458.11 ± 357.9 nanograms per gram, followed by scorpions at 457.48 ± 264.06 nanograms per gram and spiders at 340.70 ± 184.72 nanograms per gram. Variation was often substantial: for several groups, including harvestmen, flies, Vespidae wasps and ants, the standard deviation exceeded the mean. These broad spreads may reflect genuine differences among taxa, but the study’s relatively coarse taxonomic resolution, together with the analysis of whole animals without prior removal of gut contents, likely added variability.

Bioaccumulation factors were calculated by comparing mercury in each organism with its dry-weight concentration in soil or litter. A factor below 1 was classified as no significant bioaccumulation, 1 to 2 as microaccumulation, and above 2 as macroaccumulation. Overall, 83.75% of the analysed samples had a bioaccumulation factor of at least 1 relative to soil, and 91.4% of those accumulators met that threshold relative to litter. Among samples showing accumulation potential, 80.38% were macroaccumulators when compared with soil and 68.53% when compared with litter. Spiders were macroaccumulators in both comparisons, while ants showed a stronger tendency to exceed the macroaccumulation threshold relative to soil than to litter. The divergence between the two reference compartments reinforces that the surrounding soil may be an important exposure pathway for some animals.

Four mercury concentration classes were then compared with the feeding habits assigned to the organisms. Predator frequency rose from 9.1% in the lowest class, no more than 43.50 nanograms per gram, to 30.1% in the next class, 38.1% in the third and 45.5% in the highest class, above 273.81 nanograms per gram. Phytophagous organisms moved in the opposite direction, declining from 26.1% of the lowest class to 15.0%, 8.2% and 2.3% across the progressively higher classes. Omnivores predominated in the second group, while necrophagous organisms were proportionally more frequent in the highest. This shift is consistent with mercury transfer through feeding relationships, particularly the ingestion of contaminated prey, but the study characterized the inference as a local trophic-network pattern rather than a direct measurement of every trophic pathway.

Multivariate analysis provided additional support for an association between ecological function and mercury burden. The analysis combined organism mercury concentration, the bioaccumulation factor based on litter and mean body size. The first two discriminant functions explained 96.69% of the separation among feeding groups, and their profiles differed significantly at p < 0.0001. Xylophagous and sap-sucking organisms occupied similar positions, whereas detritivores, necrophages, parasitoids and predators showed partially overlapping profiles. Predatory and omnivorous groups lay closer to the centre. The overlap demonstrates that organisms with broadly similar feeding strategies did not all accumulate identical amounts of mercury, indicating that diet is not the only relevant variable. Exposure through soil contact, habitat use, body size, taxonomy and other physiological or ecological traits may also influence tissue concentrations.

The authors identified several important limitations, including the inability to identify many specimens to genus or species, large within-group variation and possible contributions from undigested material in whole-body samples. They recommended species-level identification and gastrointestinal depuration in future work. The study also took place in gallery forests inside a peri-urban reserve bordered by urban and agricultural landscapes, where previous research found elevated mercury near roads and attributed it to traffic-related emissions. Regional atmospheric deposition, wildfires and natural geological contributions are also plausible influences on the reserve’s mercury cycle. Accordingly, the findings describe bioaccumulation and patterns consistent with biomagnification in one Brazilian savanna system; they do not establish clinical effects in people, prove a specific mercury source or show that the same concentration gradients occur throughout the Cerrado. Nevertheless, the data demonstrate how litter, soil properties, organism–environment contact and feeding relationships can jointly structure terrestrial mercury transfer in a relatively under-studied soil food web.

Subject of Research: Terrestrial mercury bioaccumulation and trophic transfer in Brazilian Cerrado soil invertebrates

Article Title: Mercury dynamics in surface-active soil invertebrates from an ecological reserve in the Brazilian savanna

Article References: Fernandes, I. O., Monteiro, L. C., de Miranda, V. L., de Souza, J. P. R., Rodrigues, Y. O. S., de Castro Paes, É., Júnior, W. F. A., & Bernardi, J. V. E. (2026). Mercury dynamics in surface-active soil invertebrates from an ecological reserve in the Brazilian savanna. Environmental Monitoring and Assessment, 198(11), Article 1230. https://doi.org/10.1007/s10661-026-16051-0

Image Credits: AI Generated

DOI: 10.1007/s10661-026-16051-0

Keywords: mercury, bioaccumulation, biomagnification, soil invertebrates, Brazilian Cerrado, terrestrial food webs, leaf litter, soil contamination, ecological monitoring, dynamics, surface-active, soil