Despite growing recognition of the conservation value of grassy biomes, our understanding of how to restore biodiverse tropical and subtropical grassy biomes (grasslands and savannas; TGB) remains limited. Several tools have recently been identified for TGB restoration, including prescribed fires, appropriate management of livestock and wild herbivores, tree cutting and shrub removal, invasive species control, and the reintroduction of native grasses and forbs via seeding or transplants. However, additional research for improved TGB restoration is needed. This article aims to identify ecological research priorities for TGB restoration. The following points are crucial to scale up TGB restoration and meet the challenges of the UN Restoration Decade. Research should focus on: disentangling the reasons why TGB are often undervalued and misunderstood; mapping TGB restoration opportunities; identifying regions where TGB and other biomes naturally exist as alternative stable states; recognizing areas with natural regeneration potential to avoid unnecessary intervention; restoring soil conditions; disentangling factors driving low seed quality, determining germination requirements and developing vegetative propagation techniques for TGB species; disentangling the limiting factors and key ecological processes underlying seedling establishment and community assembly; improving and validating long‐term management to mimic natural disturbance regimes; setting the minimum attributes of desirable TGB in terms of structure, composition, functioning, and resilience; and improving monitoring of restoration outcomes. Such research has the potential to advance theory, policy, and practice in TGB restoration, ultimately resulting in long‐term benefits for people and nature in some of the more neglected ecosystems of our planet.

Aims:Indicators that can provide information during the early stages of restoration are very useful for predicting restoration outcomes. We posed the following questions: Can savanna restoration outcomes be affected by the initial functional-group composition? Are there functional groups that, when established early, can prevent colonization by invasive grasses, trigger a successful restoration trajectory, and be used as early indicators of restoration success? Methods:We established 110 plots spanning a naturally occurring range of species composition in savanna areas that had been restored through direct seeding. We looked for different initial compositions of the following functional groups: perennial grasses, annual grasses, short-lived shrubs, and invasive grasses. We measured vegetation cover over two years and evaluated the effects of the initial functional composition on the successional trajectory of the plots. Results:The initial dominant functional group determined the assembly trajectory.Short-lived and fast-growing species were replaced by perennial species, indicating a fast species turnover. Invasive grass cover remained stable over time, demonstrating that once they establish and dominate an area, an alternative stable state is achieved. Conclusions:Our results demonstrate the importance of introducing a mixture of functional groups when restoring savannas in severely disturbed areas. Fast-growing and short-lived species are important for quickly covering the ground, creating the conditions for the establishment of perennial grasses that naturally dominate the herbaceous layer of savannas. Trees are also characteristic of the savanna structure and must be introduced. Our results indicate that invasive grasses must be eliminated and/or constantly controlled before native species are introduced. K E Y W O R D Sassembly trajectory, cerrado, direct seeding, functional group, invasive grasses, priority effects, savanna restoration, species turnover, succession 62 |

Defining the reference system for restoration projects in regions characterized by complex vegetation mosaics is challenging.Here we use the Cerrado region of Brazil as an example of the importance of clearly defining multiple natural and anthropogenically altered states in grassland-savanna-forest mosaics. We define three main, natural vegetation types-grassland, savanna, and scleromorphic (cerradão) forest-to (1) distinguish between original and degraded states and (2) set appropriate targets for and guide restoration. We contend that the differences in Cerrado vegetation composition originally were driven by soil conditions and secondarily by fire frequency and precipitation patterns that differ from the core to the edge of the Cerrado region. Grasslands are found on the shallowest, least fertile soils and/or in waterlogged soils; scleromorphic forests are generally located on deeper, more fertile soils; and savannas occupy an intermediate position. In recent decades, this biophysical template has been overlain by a range of human land-use intensities that strongly affect resilience, resulting in alternative anthropogenic states. For example, areas that were originally scleromorphic forest are likely to regenerate naturally following low-or medium-intensity land use due to extensive resprouting of woody plants, whereas grassland restoration requires reintroduction of grass and forb species that do not tolerate soil disturbance and exotic grass competition. Planting trees into historic grasslands results in inappropriate restoration targets and often restoration failure. Correctly identifying original vegetation types is critical to most effectively allocate scarce restoration funding. Implications for Practice• Land managers and scientists should collaborate to identify the range of natural and anthropogenic states in grassland-savanna-forest mosaics to choose appropriate targets for restoration. • Identifying natural and anthropogenic factors influencing these vegetation types and their degraded states can help guide selection of the most suitable and cost-effective restoration techniques. • The resprouting ability of woody species allows for high resilience under low-intensity disturbance regimes; however, herbaceous native species rarely recover naturally following extensive soil disturbance and exotic grass invasion. • There is an urgent need to improve evidence-based restoration techniques in the Cerrado grassland-savannaforest complex, especially how to control invasive grasses, reestablish soil conditions, and manage fire, since techniques applicable at a large scale are necessary to achieve restoration commitments and targets.The following information may be found in the online version of this article: Figure S1. Map of Brazilian biomes. The Cerrado region (light green) is in central Brazil and borders four different biomes. Figure S2. Photos of the three main Cerrado vegetation types: (A) grassland; (B) savanna, and (C) cerradão (scleromorphic forest). Photos: Bruno M. T. Walter.

Tropical dry forest tree species are recognized for their high resprouting ability after disturbance. We tested whether species that commonly produce root and stem suckers can be propagated by large stem and root cuttings, a useful method for landscape restoration programs. We performed four experiments: (1) In a greenhouse, we tested the propagation of six species using large stem cuttings collected from early successional sites. We used the following treatments: (i) dry season collection and planting; (ii) dry season collection, storage in humid soil, and wet season planting; (iii) wet season collection and planting; and (iv) wet season collection and planting after treatment with commercial NAA auxin. (2) Stem cuttings of Myracrodruon urundeuva were planted in a pasture during the rainy season after either NAA, IBA, or no auxin treatment. (3) As a control experiment, we also planted cuttings of Spondias mombin, a species known for successfully regenerating from cuttings. (4) Root cuttings of six species were collected in recently plowed pastures and planted in the greenhouse with and without treatment with NAA auxin. No root cuttings rooted. Only M. urundeuva and Astronium fraxinifolium stem cuttings rooted. Maximum success was obtained for stem cuttings collected and planted in the dry season (23%). Only 13% of M. urundeuva had sprouted by the 15th month of the field experiment. As a result, large cuttings are not recommended for propagation of the studied species. Future studies should include development of suitable methods of root harvesting and prospection of traditional knowledge for species selection.

It has been increasingly argued that ecological restoration should focus more on targeting ecosystem services than on species composition of reference ecosystems. In this sense, the role that species play on community assembly and functioning through their functional traits is very relevant, because effect traits mediate ecosystem processes, ultimately resulting in provision of ecosystem services. One major challenge in ecological restoration is to know which species to use that will deliver the target ecosystem services. We developed an algorithm to select the minimum set of species that maximize the functional richness (FRic) and the functional redundancy (FR) of the restored community, a proxy for the provision of multiple ecosystem services and the resilience of the system to environmental changes, respectively. For this, we simulated the restoration of 24 riparian woody communities of the Brazilian Cerrado. Using the species pool of each original local community, we ran restoration simulations for gradually increasing species richness until reaching the total species richness of the original local community. We computed FRic and FR for each simulated restoration community using the traits specific leaf area, maximum plant height and seed mass. Our simulation results indicate that multiple ecosystem services could be restored with an average of 66% of the species of the original community. Moreover, an average of 59% of the species would be needed to restore communities resilient to environmental changes. Our approach contributes to solving one of the major challenges of ecological restoration, which is defining how many and which species should be used to achieve functional targets. We believe this approach can help in projects of restoration by enabling restoration practitioners to select minimum alternative sets of species that optimize the provision of multiple ecosystem services in a resilient restored ecosystem.

Upscaling trait‐based restoration to regional levels is necessary as we enter the UN Decade on Ecosystem Restoration. For this, we need to select species that achieve functional targets considering the regional species pool. Here, we present a framework to achieve multiple restoration targets using a regional species pool containing the species available on the market, species unavailable, and species that occur in reference ecosystems. The framework enables optimising functional diversity (FD), recovering FD and composition using reference ecosystems, reducing cost and increasing species diversity in restoration. Additionally, our framework allows the detection of functionally relevant species in the regional pool that are unavailable for restoration on the market. We illustrate our framework with a data set of Brazilian savanna tree communities. It was not possible to optimise FD with the species available on the market. To achieve this target, it would be necessary to use unavailable species from the regional pool. However, with the species available on the market it was possible to obtain communities resistant to fire and to restore functional composition and diversity to levels similar to or greater than those observed in reference ecosystems. Synthesis and applications. Our framework selects species to achieve multiple targets in large‐scale trait‐based restoration initiatives. The framework shows a range of solutions that can be achieved with the regional species pool. That allows the restoration practitioner to verify if functional parameters are truly optimised and which species should be added to the market or collected from the wild to achieve restoration targets. It also shows how different selected communities are from reference ecosystems, avoiding the unintentional creation of novel ecosystems.

The use of different legume species, as green manure, may affect differently soil biological and chemical properties. The aim of this study was to evaluate the effect of four legumes species used as green manure on soil biological and chemical properties in short-term. We evaluated the following legume species: Crotalaria, Cajanus, Mucuna and Canavalia. The study was arranged in a completely randomised design with four replicates. The plants were incorporated into the soil (0-20 cm) by harrow and the chemical and biological properties were evaluated 30 and 60 days from the incorporation. Soil chemical and biological properties showed different trends according to legume species used. Soil P and K contents were highest in plot with Crotalaria, while soil Ca content was highest in plot with Mucuna. Soil microbial biomass was higher in plot with Mucuna as compared with others green manure species. Fluorescein diacetate hydrolysis was higher in plots with Mucuna and Canavalia than the others plots. Our results supported the hypothesis that different types of legume used as green manure affect differently the biological and chemical properties of soil. In this case, Mucuna was more effective to improve soil biological properties, while Crotalaria seems to be more efficient in the improvement of chemical properties.

No abstract