The research team, led by R. B. Dextro and colleagues at the Centre for Nuclear Energy in Agriculture (CENA) at the University of São Paulo, selected two neighboring farms with comparable land-use histories but contrasting management strategies. AFS19, established 19 years ago, has relied predominantly on organic fertilization using manure as its principal external nutrient source, with less frequent liming. AFS27, established 27 years ago, has followed a mineral-fertilization approach for more than two decades, using chemical fertilizers and regular liming to control soil pH. Both farms grow açaí (Euterpe oleracea) and cacao (Theobroma cacao) within diversified perennial-crop systems, allowing the researchers to compare the same two crops under contrasting agroforestry conditions within the same regional landscape.

Using inductively coupled plasma optical emission spectrometry (ICP–OES) after microwave-assisted acid digestion, the team measured twelve macro- and micronutrients in freeze-dried fruit tissues. The elemental profiles were broadly similar between the two systems, but statistically significant differences emerged for specific elements. Açaí pulp from AFS19 contained 38% higher copper and 35% higher phosphorus concentrations than açaí from AFS27, while cacao seeds from AFS27 showed 17% to 30% higher concentrations of calcium, manganese, magnesium, and sulfur compared with cacao from AFS19. Potassium was the element occurring at the highest concentration in cacao samples from both systems, exceeding 5.5 grams per kilogram on a dry-weight basis, and was statistically higher in AFS19 samples. These patterns align with known soil chemistry mechanisms: organic amendments can influence phosphorus availability through changes in soil organic matter and microbial processes, whereas liming can affect calcium and magnesium availability, though the authors caution that soil nutrient availability and fertilizer application rates were not quantitatively measured.

Untargeted metabolomic profiling using liquid chromatography coupled to electrospray ionization ion-trap mass spectrometry revealed differences in the compounds tentatively annotated in samples from the two systems. Açaí extracts from AFS19 showed a greater number of putative compounds, including apigenin-7-O-glucoside, apigenin 7-O-glucosylglucoside, quercetin-3-O-rhamnoside, and caffeic acid, tentatively detected only in AFS19 samples. Quercetin and oleic acid were tentatively annotated only in AFS27 samples, while catechin, kaempferol, eriodictyol, scoparin, quercetin-3-O-glucoside, and procyanidin B1 appeared in samples from both systems. In cacao, compounds including esculin, orotidine 5-monophosphate, fructose 1,6-diphosphate, amorphigenin, tricin, and kaempferol-3,7,4′-trimethyl ether were tentatively detected only in AFS19 samples, whereas procyanidin B1, tryptophylglycine, and phenylacetyl aspartic acid were tentatively detected only in AFS27 samples. Gibberellin-related compounds were annotated in both systems, though different forms were assigned to each, and sulfojasmonate appeared in samples from both farms.

Transcriptomic analysis using RNA sequencing on an Illumina NovaSeq X Plus platform revealed distinct gene-expression patterns between the two systems. For açaí, reads were mapped against the Elaeis guineensis reference genome, the taxonomically closest palm species with a sequenced genome available, since no high-quality reference genome for Euterpe oleracea was available. Out of 29,571 annotated genes assessed, 72.9% had detectable expression levels, with only 1.3% identified as significantly different between the two systems. Hierarchical clustering based on expression profiles separated açaí samples into two principal groups, with transcripts annotated as thaumatin-like protein 1B, pathogenesis-related protein PRB1-3, oxidases, and laccases showing higher expression in several AFS19 samples, while expansin B18, non-specific lipid-transfer protein 1, anthocyanidin synthase, and chalcone synthase showed higher expression in the AFS27 group. Notably, one AFS19 sample corresponded to the white açaí phenotype, characterized by the absence or strong reduction of anthocyanin pigmentation, which clustered separately and illustrates the contribution of within-farm genotypic variation to the dataset.

The cacao transcriptomic differences were more pronounced. Out of 23,797 possible genes in the Theobroma cacao reference genome, 18,333 registered some level of expression, and 22.5% of expressed genes were significantly different between the two systems, corresponding to 4,124 statistically different genes. In AFS19 samples, genes annotated as hydroxymethylglutaryl-CoA lyase, fasciclin-like arabinogalactan protein 3, expansin B1, and chaperone protein dnaJ 11 showed higher expression, along with transcripts associated with seed protection and carbohydrate metabolism on reserve mobilization. In AFS27 samples, higher expression was observed for transcripts annotated as disease resistance protein At4g27220, LRR receptor-like serine/threonine-protein kinase FLS2, peroxidase 47, xyloglucan endotransglucosylase/hydrolase protein 9, and cysteine-rich receptor-like protein kinase 29, along with genes related to pathogen resistance, seed coating, and maturation. An endochitinase (LOC18610674) was among the highly expressed genes in AFS19 samples, belonging to a gene family that can participate in responses to wounding, osmotic stress, and pathogens.

The correspondence between elemental, metabolomic, and transcriptomic patterns provides several hypotheses for future investigation. The team observed that transcripts annotated as GA receptor GID1, GA 2-β dioxygenases, and gibberellin-regulated proteins were detected in the cacao transcriptomes, corresponding with the tentative metabolomic annotation of gibberellin-related compounds in both systems. Similarly, jasmonate- and inosine-related compounds were tentatively annotated in samples from both systems, and jasmonate signaling is known to participate in multiple plant developmental and defense processes. However, the authors emphasize that these multi-omics observations are correlative and that functional experiments, targeted chemical validation using authentic reference standards, standardized developmental sampling, and replicated environmental measurements will be required to establish mechanistic relationships among these molecular patterns.

The study’s observational design carries important limitations that the authors address directly. Because AFS19 and AFS27 each represent a single independent agroforestry system, farm identity is completely confounded with management history and also with other characteristics that may differ between the systems, including system age, vegetation structure, local environmental conditions, soil properties, crop genotype, and historical management. Trees and fruits sampled within each farm characterize within-system biological variability but do not constitute independent replicates of the respective management regimes. Sampling was additionally performed at a single time point near the end of the harvest season, and fruit maturity was not quantitatively standardized, meaning developmental differences represent a plausible explanation for part of the metabolomic and transcriptomic variation. The metabolomic analysis was exploratory, including only two tree-level samples per system, and compound annotations were not validated using analytical standards.

Despite these constraints, the integration of elemental profiling, untargeted metabolomics, and transcriptomics identified biochemical and gene-expression differences between açaí and cacao fruits sampled from the two neighboring agroforestry systems that would not have been apparent through any single analytical approach. The findings illustrate the extent to which fruits produced in real-world agroforestry settings can vary across elemental, metabolomic, and transcriptomic dimensions, and generate candidate biochemical and molecular traits for future replicated studies examining relationships among agroforestry management, environmental conditions, fruit development, and crop characteristics. As Amazonian bioeconomies continue to expand and açaí production intensifies, understanding how management practices within diversified agroforestry systems influence fruit biochemistry will be essential for optimizing both nutritional quality and ecological sustainability across tropical production landscapes.

Subject of Research: Comparative biochemical and transcriptomic profiling of açaí and cacao fruits from Eastern Amazon agroforestry systems

Article Title: Comparative biochemical and transcriptomic profiling of açaí and cacao fruits from Eastern Amazon agroforestry systems

Article References: Dextro, R. B., Virgilio, A., Alencar, S. M., Cruz, A. F., Ono, N., & Tsai, S. M. (2026). Comparative biochemical and transcriptomic profiling of açaí and cacao fruits from Eastern Amazon agroforestry systems. Frontiers in Sustainable Food Systems, 10, Article 1970644. https://doi.org/10.3389/fsufs.2026.1970644

Image Credits: AI Generated

DOI: 10.3389/fsufs.2026.1970644

Keywords: Comparative, biochemical, transcriptomic, profiling, cacao, fruits, Eastern, Amazon, agroforestry, systems, scientific research