Knowledge concerning the processes involved in defining the boundaries between rainforests (fire-sensitive) and open formations (fire-tolerant) is essential to safeguarding biodiversity and ecosystem services, especially under climate change and increased anthropogenic pressure. Here, we investigated the main environmental factors involved in the co-occurrence of forest islands and humid grasslands located in a protected area in the Espinhaço Biosphere Reserve, southeastern Brazil. We used permanent plots to collect the soil variables (moisture and chemical properties) in the forest islands. For sampling in wet grasslands, we installed four lines of 30 m from the edge of the islands in different directions. Subsequently, we delimited three points on each line10 m apart, totaling 12 points per area. We also surveyed the vegetation cover before and after prescribed burns. The environmental variables were subjected to tests of means and principal component analysis. We observed higher values of potassium, sum of bases, cation exchange capacity and organic matter in soils from forest islands than in wet grasslands. Therefore, the boundaries’ definition between the two vegetation types appeared to be primarily related to soil fertility and moisture gradients. After prescribed burning of the areas, no regeneration of arboreal individuals was detected near the edges of the islands. Therefore, our results suggest that forest islands are unable to expand due to well-defined edapho-climatic conditions. Thus, these environments should be a target focus for designing public conservation policies because they increase the complexity of the landscape of Campos Rupestres vegetation (mountain rocky grasslands).

Changes in soil nitrogen mineralization can impact nutrient availability, and further affect plant growth. It is unclear, however, how temperature elevation in alpine grassland will affect soil net N mineralization rate (N_min) across altitudes. At six altitudes (3200-4200 m with an interval of 200 m) along a slope in Lenglong mountain in the northern Qinghai-Tibetan Plateau, we performed an in situ soil incubation experiment by using the resin-core method to assess altitudinal variations of N_min. Meanwhile, we evaluated the effects of temperature elevation on N_min and its temperature sensitivity (Q₁₀) through a soil downward transplantation experiment based on three reference baseline altitudes (3800, 4000 and 4200 m). The results showed that high altitudes generally led to low values of N_min. Structural equation modeling analysis revealed that N_min along the altitude was mainly controlled by soil temperature. Increased temperature caused by the altitude transplantation significantly elevated N_min for all of the three reference altitudes. The value of Q₁₀ was 3.4 for soil samples transplanted from the reference altitude of 4200 m, which was about twice that of the lower reference altitudes of 4000 and 3800 m.

Understanding how community phylogenetic and functional structures change over succession has gained increasing attention during the last decades, but the lack of long-term time-series data has limited our understanding of the patterns and mechanisms of these changes. This is especially the case for forest communities. Here, we used an exceptionally long-term data over 68 years to analyze the secondary succession dynamics of a subtropical forest in Southeast China. We found that community phylogenetic and functional structures showed opposite temporal trends. The mean pairwise phylogenetic distance between species increased, but the mean nearest taxon distance decreased over succession, indicating both phylogenetically distantly related and sister species co-occurred in late-successional communities. In contrast, both the mean pairwise functional distance and mean nearest functional distance between species decreased over time, and community functional structure switched from overdispersion to clustering. We further distinguished the contributions of species colonization and extinction to community structural changes. We found that the new colonists were generally more distantly related to each other and to the residents than the local extinct species, resulting in increased phylogenetic overdispersion over succession. In contrast, from a functional perspective, we found that species with more similar traits to the resident species had a greater chance to colonize but a lower chance to go locally extinct, which shifted community functional structure toward clustering. Together, our study highlights the critical role of species colonization and extinction in disentangling assembly mechanisms underlying community phylogenetic and functional structures over long-term succession.

Expansion of global trade and acceleration of climate change dramatically promote plant invasions. As a result, a large number of habitats harbor multiple invasive plant species. However, patterns of invasive interactions and the drivers mediating their interactions remain unclear. In this greenhouse, potted plant study, we tested the impacts of 18 invasive plant species on the growth of target invader Erigeron canadensis which is dominant in central China. Neighboring invasive species belong to three functional groups (grass, forb and legume) and have different levels of relatedness to E. canadensis. Growth of E. canadensis’ strongly depended on the identity of neighboring invaders. Some neighboring invasive species suppressed growth of E. canadensis, others had no effect, while some promoted growth of E. canadensis. Through analyses of functional and phylogenetic similarities between the target species and neighboring invaders, we showed that two factors probably play roles in determining the relative responses of E. canadensis. Generally, E. canadensis responded negatively to invasive grasses and forbs, while it responded positively to invasive legumes. Furthermore, the negative responses to neighboring invasive grasses and forbs increased with increasing phylogenetic distance between the neighbors and E. canadensis. In contrast, the positive responses to invasive legumes did not depend on phylogenetic distance from E. canadensis. Our results suggest that successful invasion of E. canadensis probably depends on the identity of co-occurring invasive plant species. Interactions between E. canadensis and other invasive species should help managers select management priorities.