In this review, we focused on the description of plant-vegetation-ecosystem structure and soil microbial diversity (Figure. 1). We avoided the description of the reasons for which vegetation diversity and ecosystem functioning are assessed since these metrics are commonly measured in forest restoration. Then, we reported evidence of implementing the next-generation sequencing technique on environmental DNA to evaluate soil microbes, highlighting its value in restoration contexts.

This literature review aimed to assess the state of knowledge on the integrated evaluations of forest restoration actions, focusing on the relationships between the ecosystem parameters and soil microbes. We searched articles including valuations of vegetation diversity, ecosystem structure and functioning, and soil microbes, to understand if these parameters have been collectively assessed in forest restoration and how microbes change according to restoration development. Specifically, we investigated the presence of common patterns of variations of soil microbial diversity across the forest restoration chronosequence, considering the location and biome in which the research was conducted, using the biomes definition of Dinerstein et al. (2017), and disentangling general yet useful dynamics in restoration trajectory evaluation.

Our examination sought to answer the following questions: (i) What are the advances from using multiple parameters in restoration assessment (vegetation diversity, ecosystem structure and functioning, and soil microbes with eDNA analysis)? (ii) Are there general temporal and spatial patterns in microbial community characteristics (taxonomic diversity, functional diversity, and relative abundances) compared to other ecological parameters? Are these patterns consistent across different biomes or conditions? (iii) What are the gaps in integrated forest restoration assessments?

Finally, guided by the results we suggested future actions to improve the integrated evaluation of forest restoration, acknowledging the obstacles and opportunities for its implementation.

Assessments of restoration outcomes currently lack an integral and standardized methodology, limiting their comparability and robustness. Several indexes and protocols have been implemented to measure vegetation diversity, its structure, soil fertility, and microbial community characteristics (Ruiz-Jaén and Aide, 2005b; Bartholomew et al., 2024). Each metric emphasizes different aspects of the parameter assessed, making direct comparisons impossible. For instance, both Shannon and Chao1 indices measure diversity, but the former calculates richness and diversity considering the observed relative abundance of taxa, while the latter evaluates estimated species richness (Chao, 1984; Shannon, 1949). Furthermore, multiple techniques are employed to perform microbial DNA extraction and sequencing (Hakimzadeh et al., 2024), limiting the comparison of results across studies and locations. In addition, the lack of standardized and generally adopted measures of dead wood presence imposes a significant limit on understanding the presence and dynamics of identified microbial taxa. Some taxa rely entirely on dead wood material for their survival (Baldrian et al., 2016), making this a fundamental information to reliably compare microbial communities across locations and stages of restoration.

Analyzing distinct restoration metrics implies trade-offs. The diverse number and type of metrics among studies impose simplifications, which can lower the quality of the resulting analysis, but ensure a shallow yet sufficient contrast, assuming the same robustness and reliability of the results for each study. General comparison between studies can still be accomplished because, despite the divergent metrics adopted, the overall aim of assessing restoration outcomes is a shared goal. For example, advances in the assessment of restoration dynamics from microbial analysis using the Shannon diversity index, and canopy coverage for vegetational structure, may seem comparable only to studies utilizing the same metrics. Nevertheless, comparing research exploiting a different diversity index (e.g., Pielou index), and average DBH as a structure metric is possible since the overall aim is to estimate the advances in integrated restoration evaluation (as in this review) and not the effect of the techniques and indexes adopted.

From a total of 175 articles, we retained 30 for this review (Supplementary Table S1). Approximately half (57%, 17 studies) of the retained articles evaluated multiple parameters: vegetation diversity, soil microbe characteristics, ecosystem functions such as soil organic carbon (SOC), and forest structure. Forty percent (12 studies) considered vegetation diversity, soil microbe characteristics, and ecosystem functions, and the remaining 7% (2 studies) examined the vegetation diversity, soil microbe characteristics. Seven of the articles (23%) used a list of dominant species as a metric for vegetational diversity, while the others included a more comprehensive list of species. Of the 30 studies, 30% (9) considered both bacteria and fungi, 30% (9) analyzed only bacteria, and 30% (9) solely studied fungi. For the last three studies, one considered the kingdom of Archaea in combination with bacteria and fungi, while in the remaining 2 articles, the analysis regarded only diazotroph bacteria and Archaea, respectively.

Ecosystem functions were more commonly assessed than ecosystem structure. Ninety-seven per cent of the articles (29 studies) evaluated at least one ecosystem function such as SOC, total carbon, and soil organic matter, and 76% (23) calculated SOC in combination with other soil parameters (e.g., pH, total nitrogen), with two studies also determining the decomposition rate of organic matter in the soil. Fifty-six per cent (17) of the retained articles included the per cent cover, crown, and canopy density. Tree DBH data was collected in 36% (11) of the articles studying forest structure. Field data collection for all the articles occurred between 2010 and 2021, with 23% (7) of the studies conducting fieldwork in 2019, and 20% (6) in 2015. The average age of the restored, reforested, or revegetated plot was 29.5 years, with three studies surveying variations within 5 years of project implementation.

Metrics used to assess taxa richness included Operational Taxonomic Units (OTUs, 28 studies), and Amplicon Sequence Variants (ASVs, 2 studies) (Fernandez Nuñez et al., 2021; Gómez-Aparicio et al., 2022). Calculation of ASVs is considered to be more accurate than OTUs for separating and clustering microbial groups based on phylogeny, but the methodology is congruent at the beta-diversity level, allowing the comparison of studies in this review (Chiarello et al., 2022). Indexes used to evaluate microbial richness and diversity were Shannon, Simpson, Pielou, ACE, Faith (PD), and Chao1. See Supplementary Table S2 and discussion by Kim et al. (2017) for a comparison between these diversity indexes.

Functional groups of bacteria and fungi were assessed in 22 of the 30 articles analyzed. The assessment of functional groups in these studies was based on the association of the evaluated taxa to the presence of genes encoding for specific functions. Soil enzyme activities were explored in one article (Huang et al., 2023). Sequencing of functional genes was performed in one study (Mirza et al., 2020), and a MicroResp system was used in two articles to characterize the metabolic profiling of microbes (Bonner et al., 2019; Smenderovac et al., 2017). Metabolic processes were deducted in two studies for bacteria soil communities (Deng et al., 2019; Li et al., 2023). The remaining articles assessed microbial functional groups using the following databases: FAPROTAX, FUNGuild, Fungal-Traits, PICRUSt2, and Index Fungorum.

Of the 30 articles evaluated, 17 (56%) collectively considered the vegetation diversity, ecosystem structure and functioning, and soil microbes studied through eDNA. Fourteen of these studies (82%) were conducted in China. Analyzed biomes, as defined by Dinerstein et al. (2017), included boreal forests/taiga, Mediterranean forests, woodlands and scrub, temperate broadleaf and mixed forests, temperate conifer forests, tropical and subtropical dry broadleaf forests, and tropical and subtropical moist broadleaf forests (Figure 2). Respectively, 100% (9) and 75% (9) of the studies conducted in the temperate broadleaf and mixed forests, and in tropical and subtropical moist broadleaf forests occurred in China.

Bacteria community composition became increasingly similar to remnant forest over vegetation succession in the Mediterranean and tropical biome. Bacterial taxonomic diversity and soil chemical characteristics remained the same in a Mediterranean biome over a successional chronosequence of 10 years, while bacterial taxa relative abundances over time resembled those in remnant reference sites (Gellie et al., 2017). The peak of diversity at this site occurred in the 7th year, whereas the community composition (taxa relative abundance) was increasingly similar to that of remnant sites over time (Yan et al., 2020). In this case, the soil physicochemistry was correlated with the age of the restoration. For example, nitrate and phosphorus were negatively correlated with time since restoration and ammonium and organic carbon significantly increased over the 10 years (Yan et al., 2020). In tropical and subtropical moist broadleaf forests, soil diazotrophs bacteria community composition increasingly converged with a primary forest over 40 years. Bacterial community composition also varied along with soil fertility (e.g., higher C content occurred in the pasture soil compared to primary forest sites) and was closely associated with plant species variation (Mirza et al., 2020).

Bacteria diversity showed discordant development patterns over the forest succession in temperate areas. In a temperate broadleaf and mixed forest biome in the Loess Plateau (China), bacteria showed higher diversity accompanied by higher soil C values in natural and artificial restoration plots (78 and 42 years old respectively) compared to plots of crops (Yang et al., 2018). In an additional study conducted in China, bacteria demonstrated an increase in diversity with vegetation restoration, and the soil chemical properties became comparable to the ones present in the reference site (Li et al., 2023). In 40-year-old mixed coniferous—broad-leaved plantations, coniferous plantations, and broad-leaved only plantations in North China, bacteria diversity and vegetation community structure were higher as the plantations got older compared to the early successional stage where shrubs dominated the landscape. Soil chemical properties also showed a differentiation between the early phases of succession and the forested stages (Qiu et al., 2022). Bacteria diversity decreased over a restoration period of 30 years of Pinus sylvestris and P. tabuliformis plantations in China, and soil carbon content increased during forest restoration (Deng et al., 2019). Similar results were found by Zhang et al. (2019), with decreasing bacterial diversity over a restoration period of 150 years with Querqus wutaishanica trees (in the Shandong Province, China) (Deng et al., 2019).

Our search revealed predominantly hump-back responses of fungal taxonomic diversity variations across biomes and time scales (Table 1). Research conducted in Mediterranean and tropical biome indicated that fungal diversity shows a hump-back pattern, and that saprophytic taxa are replaced by ectomycorrhizal ones over the forest development. In the Mediterranean biome, the fungal diversity reached a peak of diversity at the 8th year, and the early dominance of saprotrophic taxa shifted toward ectomycorrhizal taxa over a chronosequence (Yan et al., 2018). In the same biome, soil chemical properties showed a directional change toward the reference state, with increasing SOC and decreasing nitrate and phosphorus. In a tropical rain forest of Costa Rica, naturally regenerating plots of mixed species of different ages (from 5 to 30 years old and mature stage forest), displayed a fungal richness peak in the mid-successional stages (15–30 years), due to the presence of both early- and late-successional species (Adamo et al., 2021). In this study, researchers found that soil Mn concentration increased over the chronosequence, while pH, organic matter, N, and P showed no changes. In the above-ground sphere, tree richness significantly increased over the successional chronosequence (Adamo et al., 2021). Taxonomic diversity of plant pathogens, wood decomposers, and root-associated fungi did not differ over the sequence, while saprophytic fungi reached a peak in the mid-successional stage. Parasitic fungal taxonomic diversity increased over succession. In another moist tropical biome study of P. yunnanensis plantations up to 80 years, fungal diversity also assumed a “hump-back shape,” reaching the highest value at 60 years (Li S. et al., 2020). The hump-back-shaped function was also observed in woody species richness. Soil chemical variables (SOC, N, P) increased over the reforestation. Ectomycorrhizal fungi abundance initially decreased and then increased with the reforestation progression.

The relative abundance of endophytic fungi also improved with reforestation, and the opposite occurred for the relative abundance of saprotrophic fungi. In 40–50-year-old plantations in tropical and subtropical moist biome Zhang et al. (2023) observed that fungal diversity declined with age. The transition from the grassland-shrub stage to the forested phase was associated with an increase in symbiotrophic fungi dominated by ectomycorrhiza and a decrease in saprotrophs. In another study, fungal diversity increased over the chronosequence in the moist tropical biome. The study, established in 1989, was undertaken in shrublands revegetated with different tree species: P. massoniana, Betula luminifera, and these two species as a mixture. The P. massoniana and mixed species forest (B. luminifera and P. massoniana), had the highest value of SOC, soil water content, total nitrogen, and total phosphorus, while pH value was comparable across stages of reforestation. The highest fungal taxonomic diversity was observed in mixed forest plots, and it was lower, but similar for the two monospecific forests. In both mixed and monospecific forests, fungal diversity was augmented with the increased age of reforestation (Liu et al., 2022). Still, the fungal taxonomic diversity and ectomycorrhizal relative abundance increased across a 30-year chronosequence, while bacteria taxonomic diversity and saprotrophic fungi relative abundance decreased in the New Caledonia tropical biome (Fernandez Nuñez et al., 2021). The decrease in saprotrophic: ectomycorrhizal fungi ratio indicated the restoration successional trajectory. SOC did not vary across the chronosequence, whereas total nitrogen decreased, and ammonium showed a peak in the mid-stage of forest restoration (Fernandez Nuñez et al., 2021).

Fungi in temperate biome plantations can display different patterns of variation. In the temperate conifer forest in Hokkaido Island Japan, fungal richness declined in monoculture plantations of Picea glehnii (a native species) and did not vary significantly over 49 years in Abies sachalinensis (also a native species) plantations or naturally regenerated forests (Naka et al., 2023). Ectomycorrhizal and saprophytic fungi recovered to the reference state of the natural forest in unmanaged regenerated forests, due to an increase of ectomycorrhizal fungi and a decrease of saprotrophic ones over the chronosequence (Naka et al., 2023).

The decline in the number of trees sporadically affected the taxonomic richness of bacteria and fungi in the Mediterranean biome but not in the Boreal one. We retained two studies considering the decline of tree numbers to better understand the patterns of richness and diversity variation of fungi and bacteria. In a Jack pine plantation in Canada (boreal forest/taiga biome), the number of trees diminished due to harvesting, and this action did not impact bacteria, Archaea, and fungi taxonomic richness. However, the decline negatively affected the microbial metabolic activity (Smenderovac et al., 2017). In a study conducted in the Mediterranean biome, the ASVs variation displayed an increase in bacteria and fungi diversity with a decrease in the number of trees caused by a pathogen invasion 25 years earlier (Gómez-Aparicio et al., 2022). Diversity and abundance of ectomycorrhiza declined with the reduced tree numbers, and saprotrophic fungi increased, both in diversity and abundance (Gómez-Aparicio et al., 2022).

Bacterial and fungal taxonomic diversity can display different patterns of co-variation. An afforestation study conducted in grasslands of the Loess Plateau (China), demonstrated that after 15 years of P. tabuliformis plantation, bacterial taxonomic diversity did not vary significantly, whereas fungi taxonomic diversity decreased (Wang et al., 2019). At this location, SOC decreased, along with soil nitrogen content. The relative abundance of ectomycorrhizal fungi increased, and the relative abundance of biotrophic fungi decreased (Wang et al., 2019). In another afforestation plantation of P. tabuliformis in Shaanxi Province (China), bacterial taxonomic diversity decreased over 60 years, whilst fungi taxonomic richness and SOC increased. Fungal and bacterial taxonomical community composition varied according to the succession stage (Liu et al., 2019). In P. sylvestris plantations (species native to the study location), bacterial communities had the highest diversity in one of the latest assessed stages (at age 41 of 59 years), and fungal diversity decreased until the age of 41 (Bi et al., 2021). Community composition changed, SOC initially decreased and then returned to the young stand values (15 and 30 years old). Available nitrogen and total nitrogen increased and then diminished to reach a value lower than that observed in the earlier stages. Available phosphorus significantly decreased with age (Bi et al., 2021). In Robinia pseudoacacia plantations (a nitrogen-fixing tree species with a symbiotic relationship with Rhizobium bacteria) on the Loess Plateau, taxonomic diversity of bacteria and fungi, SOC, and ammonium (NH₄⁺) were positively correlated with stand age over 45 years. Bacteria functional traits indicated a shift from oligotrophs to copiotrophs as soils transitioned from nutrient-poor to nutrient-rich (Zhou et al., 2022).

In tropical forests, the vegetation structure and composition affect the bacteria and fungi community. In a study focused on reforestation with P. massoniana (over 30 years), the presence of an understory layer was associated with higher bacteria and fungi taxonomic diversity while being positively correlated with nitrogen-fixing bacteria and negatively correlated with ectomycorrhizal fungi. Taxonomic diversity of bacteria and fungi was also correlated with SOC, nitrogen, and phosphorus, and the community composition varied according to ferns’ presence and the successional stage (Lu et al., 2023).

Bacteria and fungi diversity displayed different patterns and relationships with ecosystem function in native and exotic tropical plantations. In native plantations of P. massoniana, Cunninghamia lanceolata and Schima wallichii, Castanopsis hystrix, Michelia macclurei and Cinnamomum burmannii in Southern China, bacteria and fungi biomass and Shannon diversity were lower than in exotic monocultures of Acacia mangium, but their richness was higher. The authors highlight the possible role of the tree species attributes in shaping the microbial community diversity (e.g., lower substrate quality furnished by the native conifer species for determined microbial group). A significant relationship between soil chemical compounds was found. Above- and below-ground composition was correlated for bacteria in sites where restoration occurred with native trees. Nitrogen and SOC stocks were similar among exotic and native forests, while the total phosphorus was higher in native plantations (Wei et al., 2023).

In the Philippines, 20 years old Swietenia macrophylla plantations displayed variation in microbial community functions (substrate use, respiration, and enzyme activity) convergent to the remnant reference state (Bonner et al., 2019). Microbial community composition was further along the path of recovery in 20-year-old “rainforestation” plantations (i.e., plantations of highly diverse and mainly native species) than in monocultures of S. macrophylla. The SOC was significantly higher in the monoculture plantations and was better explained by microbial composition variation than the land-use.

Despite some advances in assessment using multiple parameters (vegetation diversity, ecosystem structure, ecosystem functioning, and soil microbes’ assessment with eDNA analysis), our review revealed a geographical gap in the application of these studies. We retrieved a list of less than 200 studies, with the first article published in 2018, substantially later than the publication of articles considering general aspects of forest restoration. Integrated evaluation in forest restoration outcomes is at the frontier stage, as is eDNA analysis. The first analysis based on environmental DNA was reported by Ogram et al. in the late 1980s (Ogram et al., 1987).

Our review highlights the lack of studies in different areas and the initial stages of forest succession. We confirmed the geographical gap found by other authors (i.e., lack of studies in less developed countries) (Ruiz-Jaén and Aide, 2005a; Wortley et al., 2013), extending it further to include North American locations that Ruiz-Jaén and Aide (2005a), and Wortley et al. (2013) evaluated as being properly assessed. Tropical and subtropical coniferous forests also lacked integrated evaluation, and a bias of studies was represented by much of the research being conducted in China, particularly on the Loess Plateau. While the scarcity of studies and biases toward geographical region may prevent the identification of other microbial variation patterns, the large occurrence of studies based in China did not hinder the research of a general microbial pattern of variation, since these analyses considered multiple biomes (temperate and tropical forests) present within the borders of this state and in other parts of the world. Restoration success is mostly evaluated after a long time (the average age of examined plots was 29 years) highlighting the absence of analyses for the early phases of forest restoration.

Bacteria and fungi were not always studied together, despite their reciprocal influences on their diversity and community composition. Study objectives, authors’ experience, and the costs of the primers and machines used influenced the selection of one of the two microbial groups. Archaea were the least investigated, despite their importance in ecosystem dynamics. While bacteria genetic libraries and ecology are becoming well-established, effort is required to build libraries for fungi and Archaea. Given this limitation, we suggest analysis to acquire baselines for these latest groups, concurrently improving bacteria’s available information, for which progress is expected for the exploitation of their metabolic activities in particular. In parallel to the enhancement of fungi and Archaea databases, bacteria may be used as a common metric to assess restoration success.

Soil characteristics were usually examined, while ecosystem structure was generally neglected. Most articles observed soil characteristics (93%), allowing possible comparison among different locations. Interestingly, one article measured the organic matter decomposition, which may lead to fruitful information about ecosystem functioning (Philpott et al., 2018). Ecosystem structure was mostly overlooked, despite its importance for microbial community dynamics in restoration contexts (Lu et al., 2023). Although succession stage and forest age may be directly related to vegetation structure, careful consideration is needed on the metrics used. For example, diameter at breast height (DBH) increases over a chronosequence but Deng et al. (2019), and Zhang et al. (2019) found that stand density decreases over this time. The explicit incorporation of measures of ecosystem structure in environmental dynamics analyses must be acknowledged, given the simple ways to acquire this data and the importance of it in restoration success and microbial variation. Evidence of the direct influence of ecosystem structure on soil microbes was provided by Fernandez Nuñez et al. (2021) and Lu et al. (2023), and it will be discussed in the following paragraphs.

The infancy of integrated restoration evaluations hinders a comprehensive assessment of trends across restoration ages and biomes, but early results suggest the inclusion of different parameter analyses, the usage of innovative methods, and coordinated efforts between stakeholders to assess the restoration outcomes and plan future actions (Table 2). For instance, while data on vegetation and ecosystem properties are often collected through time as occurred in research projects conducted in Costa Rica (Aguilar-Arias et al., 2012; Morales-Salazar et al., 2013), we suggest the inclusion of soil microbial community monitoring to better understand restoration ecosystem dynamics. The monitoring of restoration project success requires technical improvements, additional studies to fill geographical gaps, and clarification of microbial community patterns. Furthermore, divergent trends regarding the extent and type of soil microbes’ taxonomic diversity variation along restoration gradients and locations corroborate the need for research in different geographies, while developing a standardized protocol that can be applied in dissimilar ecosystems, producing comparable data (thus overcoming the related challenges stated in section 1.3).

As a standardized protocol, we highly recommend the assessment of restoration through the assessment of microbial community characteristics (taxonomic and functional diversity assessed through eDNA sequencing); evaluation of vegetation layers (coverage and taxonomy and/or functional diversity); characterization of the spatial disposition of biotic and abiotic components, tree density (stems/ha) and occurrence of abiotic features (e.g., rocks, streams, faults). For this latest evaluation, the advantage of remote sensing and laser scanning techniques to accurately monitor vegetation cover and topography (Buckley et al., 2023) has great potential.

Restoration assessment integrating plant diversity, ecosystem structure, soil microbial community, and ecosystem function metrics can significantly improve the outcome of restoration actions. Innovative methods enhance efficiency by streamlining the aggregation of multiple parameters. Remote sensing, and eDNA analysis, are great examples of innovative and cost-effective techniques that will bring environmental study to a new insightful level. Additionally, advanced practices to conduct ecosystem assessments are being produced (e.g., eDNA protocols for sample collection), as well as pioneering statistical analyses and increasing the processing power of computers (reaching the quantum field) allowing the use of cutting-edge machine learning algorithms. Examples of these are the structural equation models found in the analyses of Zhou et al. (2022), deep learning methods to investigate metabarcoding data (Lamperti et al., 2023), and revolutionary classification functions (Novello and Tonello, 2024). Plant diversity and SOC are metrics commonly implemented in restoration evaluation, but the inclusion of analyses on ecosystem structure and soil microbes’ diversity is yet to be established. Recognizing relationships and expected trajectories of all these environmental characteristics will allow stakeholders to understand what to aim for, the constraints and the best strategy to adopt to achieve the restoration aims and how they can measure the progress toward them.

The understanding of microbial community variation along with ecosystem structure yields accurate results and better hypotheses for the recognition of successional dynamics (Lu et al., 2023). The most used measure of ecosystem structure is ground cover percentage of herbaceous vegetation, and canopy cover for trees, not always in combination with DBH and other metrics. We suggest the holistic development and utilization of additional ecosystem structure parameters that can better inform restoration ecosystem dynamics.

Preliminary studies on microbial diversity may be useful to detect environmental variation and inform, further study on fungi or bacteria according to the time scale of restoration assessment. Assessing the microbial taxonomic diversity variations through time can demonstrate the successional “phase” of the ecosystem when it is not clear from the morphology or vegetational species composition. Meanwhile, the functional community composition may indicate the recovery toward the reference state (Fernandez Nuñez et al., 2021). In addition, by acknowledging the different times required by bacteria and fungi taxonomic diversity to significantly change, restoration assessments can be designed that consider trade-offs between costs and comprehensive analysis. For example, when the assessment is done in the early successional stages and the budget does not allow the investigation of different microbial groups, fungi should be the preferred focus. The opposite may apply to later-stage evaluations.

Constructive results for forest restoration management will arise through directed research on its ecological dynamics. The recent rise in the number of articles focused on integrated restoration measurements permits speculation on a future escalation of efforts to better understand ecosystem dynamics in restoration projects. The upsurge of integrated studies has been aided by the broadening of the eDNA methodology usage and the acknowledgement of the importance of soil and its microbial community.

Studies focused on different patterns of variations in monoculture and mixed plantations (e.g., if the number of ASVs reaches the same value between plantation types) are needed to understand the implications of plantation types on the ecosystem. For example, if after restoration the ASVs values are similar to the previous state, the success of the restoration could be questioned, enabling the understanding of the legacy effect of the pre-restoration biome.

Research unraveling the relationships between soil microbes and ecosystem characteristics in restoration contexts will lead to the following outcomes: (i) advances in the comprehension of relationships between ecosystem characteristics; (ii) recognition of common patterns of microbial diversity and restoration success; (iii) acknowledgement of ecosystem parameters values in determining the presence of soil microbes able to enhance restoration success; (iv) development of strategies to cope with projected environmental variations due to climate change, given the comparison of the results from different places (e.g., constraints leading to restoration success in dry locations may be applied to areas facing increased aridity due to climatic changes); (v) increase of information on fungi and bacteria presence, identity, and ecology; (vi) detections of strains of microbes that may be used for inoculation to improve plant survival and growth (Fernandez Nuñez et al., 2021).

We advocate the importance of collaboration and knowledge sharing between research entities, governments, and other actors involved in restoration programs. The establishment of an international research network would promote the application of restoration strategies in different climatic, social, and land use settings, learning from mistakes, and taking advantage of successful approaches. This should lead to a more coordinated effort toward restoration projects’ success, preventing delays from avoidable mistakes.