Historical sheep farming in the Patagonian drylands has led to reduced grass cover, soil erosion, and shrub encroachment, compromising ecosystem function. Effective restoration requires managing shrub cover, bare soil, and patch connectivity through various strategies. This study evaluates rehabilitation interventions in a grass-steppe ecosystem, comparing grazed and ungrazed areas. Over three years, we tested the following: (a) mechanical shrub cutting with biomass redistribution, and (b) enhancing patch connectivity with Pinus spp. branch piles, alongside controls, in eighteen 5 m × 5 m plots invaded by Mulinum spinosum. Half of the plots were fenced to exclude grazing, resulting in six treatment combinations. We monitored soil properties, vegetation cover, and species composition. The treatments explained twice as much of the variation in community composition as the annual climatic variations (0.26 vs. 0.13). Livestock exclusion increased perennial grass cover more than the grazed plots did (2.14 vs. 1.42 times the initial measure). All treatments reduced the amount of bare soil except the grazed controls. Shrub cutting, especially with grazing, increased the lasting litter coverage by 5–10% and decreased the bare soil equivalently. Organic matter increased except in the non-intervened interpatches (0.95 times). The enclosures with cut shrubs trapped erodible particles, showing a 5% increase. Our study highlights that grazing destabilizes communities, while enclosures stabilize them, with interventions improving soil fertility and mitigating erosion.
Fungal communities inhabiting leaves are key players in ecosystem processes but remain largely unexplored in Southern Hemisphere temperate forests. We characterized the foliar mycobiota of Nothofagus pumilio, a dominant deciduous tree in Patagonian forests, using ITS1 metabarcoding across seasons and tree health conditions. We detected 426 fungal taxa, including a 40-Amplicon Sequence Variant (ASV) core mycobiome persisting year-round. Fungal richness and biomass increased significantly in autumn, coinciding with leaf senescence, and community composition shifted markedly between seasons. Spring leaves were enriched in pathogens and basidiomycetous yeasts, while autumn leaves hosted more saprotrophs, ascomycetous yeasts, and lichen-associated fungi. Tree health had limited influence on overall community structure, but symptomatic trees showed higher ASV richness and specific indicator taxa, including the pathogen Trichosporiella multisporum and members of the Taphrinaceae and Saccotheciaceae families. Despite taxonomic turnover, ecological guilds remained relatively stable, suggesting functional redundancy. These findings reveal a seasonal successional trajectory in the foliar mycobiota of N. pumilio, from early-colonizing endophytes in spring to diverse decomposer assemblages in autumn. This study provides the first high-throughput insight into the structure and dynamics of foliar fungal communities in Southern Hemisphere temperate forests, offering a baseline for understanding microbial roles in forest health and resilience.
High-throughput sequencing approaches have become frequent in the study of endophyte communities allowing the cumulative description of fungal diversity in the last decade. However, they brought new challenges to researchers in terms of programming and developing of informatics tools. Currently, there is no consensus concerning the appropriate bioinformatics to process such sequence data. The aim of this study was to compare the performance of three pipelines of two cost-free toolkits designed to be friendly to non-programmer users, and specifically developed for fungal data: AMPtk and PIPITS. The sapwood-inhabiting fungal assemblages of two Nothofagus species from the Patagonian Forests were assessed through metabarcoding of the internal transcribed spacer (ITS) and compared with an extant sequence dataset obtained from culture prospection in the same study sites and trees. The AMPtk toolkit has performed better concerning community description in terms of precision of taxa clustering, mainly due to the DADA2 algorithm; PIPITS evidenced a higher sensitivity in detecting taxa known to be present, hence it is potentially useful for future specific taxa detection surveys. Because of a current lack of information of the reference databases, both bioinformatic toolkits performed poorly as to taxonomy assignment. It is imperative to continue studying these ecosystems to, concomitantly, improve databases and the explanatory potential of the new technologies.