Seasonal fungal signatures across Amazonian floodplain and upland soils

The Amazon rainforest represents the largest continuous tropical forest on Earth and plays a central role in global biodiversity and carbon cycling. Among its major forest types, seasonally flooded and upland forests differ markedly in hydrology, soil properties, and ecological dynamics, yet the diversity and predicted functional potential of their soil fungal communities remain poorly understood. Here, we investigated fungal community composition, diversity, and predicted functional potential in soils from two Amazonian floodplain forests and one upland forest across wet and dry seasons. Using high-throughput sequencing of the Internal Transcribed Spacer (ITS) region combined with multivariate and machine-learning approaches, we found that fungal community structure differed markedly among sites and was strongly associated with variation in soil physicochemical properties. Fungal beta diversity was primarily site-specific, reflecting pronounced environmental differences among the studied soils, whereas seasonal patterns were more evident in floodplain alpha diversity. Floodplain soils also harbored higher proportions of specialist taxa, while upland forest soils showed greater evenness, broader predicted functional diversity, and a more balanced distribution of specialists and generalists. Predicted functional guild composition also varied among sites and showed a significant site × season interaction. During the wet season, no statistically significant difference in predicted functional guild composition was detected between the two floodplain sites, whereas both remained distinct from the upland forest. Together, these findings reveal strong spatial structuring of Amazonian soil fungal communities along local edaphic gradients, while seasonal variation is more prominently reflected in diversity patterns, network organization, and predicted functional guild composition within floodplain soils.