The global demand and enthusiasm for forest restoration has never been so high (Sacco et al., 2021). The world is witnessing a rally of forest restoration and tree planting pledges, with multi-million hectares and trillion-trees campaigns (Holl and Brancalion, 2020; Goymer, 2018). But implementing large-scale, long-lasting forest restoration requires more than aspirations; it must overcome critical and unresolved challenges that have constrained restoration programs worldwide (Fagan et al., 2020; Reid et al., 2018). One of the main challenges is its financial viability. Forest restoration relies, in most cases, on the reconversion of agropastoral land uses back into forests and requires large investments to kickstart forest succession in severely degraded sites through tree-planting approaches (Holl and Aide, 2011). Incentive mechanisms, such as payments for ecosystem services, may not be enough to overcome land opportunity, implementation, and maintenance costs. A promising opportunity to make active restoration financially viable is the production of native tree species timber. Still, the lack of reliable growth data for most species has been a central limitation for timber-oriented restoration projects.

Biodiverse plantations of native species have been primarily established with the main goal of ecologically restoring native forests, maximizing habitat for multiple species (Despot-Belmonte et al., 2017), soil protection (Deng et al., 2012; Krainovic et al., 2020), hydrological services (Ahmad et al., 2001), climate change mitigation (Bastin et al., 2019; Griscom et al., 2017), socioeconomic benefits (Erbaugh and Oldekop, 2018), and legal compliance (Rodrigues et al., 2009). Conversely, the illegal exploitation of forest remnants, along with commercial monocultures of exotic species, has been meeting a portion of the market demand for timber (Brancalion et al., 2018). In Brazil, highly diverse restoration plantations have been historically established on private lands, mainly as riparian buffers, to comply with the federal legislation, whereas commercial tree plantations (9.55 Mha) are dominated by eucalypt (7.47 Mha, 78%), and pine (1.70 Mha, 18%) stands. The few commercial native tree species cultivated largely for timber in their original domain are Araucaria angustifolia (Bert.) Küntze and Schizolobium parahyba var. amazonicum (Huber ex Ducke) Barneby, which occupy, together with other native species, only 3.4% of the tree plantation area (Brazilian tree industry, 2021).

Native and exotic timber production has occurred in contrasting conditions. Whereas native timber exploitation has been concentrated in large tracts of old-growth forests in the Amazon (Brancalion et al., 2018), based on the harvesting of old, huge trees, exotic timber production has predominated in short-rotation industrial plantations located in southern Brazil, in highly degraded and fragmented landscapes within Atlantic Forest and Cerrado areas (Gonçalves et al., 2013). The timber-derived products correspond to ∼90% of Brazil's total forest market value (Brazilian tree industry, 2021). The wood exploited from native forests yielded 518 million dollars in 2020 (IBGE, 2020), while those exploited from commercial plantations yielded 3.48 billion dollars. Improving native timber and non-timber production in restoration plantations can leverage large-scale restoration (Hua et al., 2022; Lamb et al., 2005).

The competition with illegal logging has always been a severe limitation for native timber production in commercial plantations. Still, the growing legal demands for restoring degraded lands in the country may change this scenario. Under the Native Vegetation Protection Law, farmers are required to maintain a minimum of 20%, 50%, or 80% of their landholding covered with native ecosystems, depending on the region. If they do not comply with an area of Legal Reserve (LR), they must compensate it off-farm or restore it locally. Harvesting timber through low-impact logging in the LR is allowed, although this is not possible in the Atlantic Forest domain (Brasil, 2006). As the deficit of LR encompasses several million hectares (Freitas et al., 2017; Metzger et al., 2019), many landowners are obliged to restore native forests, but ideally, they may count on native timber production to offset costs and make attractive profits.

Many emblematic species with high market value have been historically exploited in natural areas and now rely on plantations to be commercialized again, but the viability of timber production in restoration projects requires growing trees of market-valued timber in a reasonable time, with straight boles and a suitable diameter for a sawmill. However, little is known about the growth potential of native timber species, especially in biodiverse restoration plantations, which are not adequately managed for timber production (e.g., use of improved genetic materials, thinning and pruning, pest- and pathogen-control). Understanding the potential of restoration plantations to produce marketable native timber is a critical step for offsetting restoration costs and transforming it into a profitable activity. Given that restoration plantations are established with multiple native tree species and usually have an abundant regeneration of native colonizing species in the understory a few years after plantation, we expect that harvesting a share of the planted species could delay, but not hamper restoration trajectory over time. The level of impact will be affected by the harvested volume, harvesting method, and site features (e.g., slope, rainfall), so integrating timber production with forest restoration relies on careful planning and monitoring.

Growth predictions are essential prerequisites for guiding silvicultural management and harvesting decisions. Forest restoration growth models represent an important tool to assess critical aspects of timber production sustainability (Günter et al., 2011), revealing opportunities to guide management interventions and reduce the pressure of exploitation on native forests.

In this study, we aimed to understand the potential for native timber production in tropical forest restoration plantations. More specifically, we built growth models, defined harvest time, and designed an optimized scenario for timber production of native tree species introduced in biodiverse restoration plantations. We also evaluated the bole quality for saw wood production.

Critical questions related to the production of native timber in restoration plantations, like "Which species to plant?", "When can I harvest some trees?", "When does the payback begin? "How much should be invested?" and "Which yield to expect?" are still unsolved, undermining investments in this activity. We explored some of these questions using a chronosequence of commercially valuable native trees from restoration plantations. Most tree species analyzed have reasonable stem quality but slow growth, at least when compared to exotic species commonly used in the tropics, such as Australian Eucalyptus (Stape et al., 2010). We discuss alternatives to minimize some of these limitations and increase the viability of timber production in restoration plantations.

Most tropical trees valued for their timber are late-successional hardwood species, which grow slowly and only reach large sizes after decades. Their initial development in the forest understory, where light availability is lower and competition is high, contributes to concentrating biomass allocation to a single stem, thus resulting in boles of great quality for saw wood production. Such characteristics have favored the exploitation of old trees in forest remnants, allowing successive harvesting cycles (Schulze et al., 2008). Restoration plantations offer contrasting conditions for timber production. Tree seedlings are planted in open areas and have high light availability during their initial development, which favors the production of multiple stems. Such conditions could also favor tree growth, but late-successional timber species compete with pioneer trees in the early stages of restoration development, severely limiting their growth (Brancalion et al., 2020; Holl and Brancalion, 2020).

Thinning and pruning would help resolve these limitations, but these techniques are not commonly used in biodiverse restoration plantations. Biodiverse stands require ecological and silvicultural knowledge about the species being cultivated for prescribing thinning and ensuring that trees receive adequate light for their full growth (Rolim et al., 2019), while experiments with well-known species have shown that the stand structure (e.g., spacing between trees), has a significant impact on wood growth (Stape et al., 2010). Tree growth limitations have been observed in reduced-impact logging operations in the Brazilian Amazon, where the slow growth of late-successional species planted using the enrichment technique has frustrated time expectations of sustainable timber management (Pinto et al., 2021). Thus, just planting commercially viable native trees in restoration areas and waiting does not seem a viable option for efficient timber production. Timber production in restoration plantations is a novelty that requires further studies and advances to become a competitive land use.

Unlocking the high-quality timber potential of native tree species may require a robust research and development program to resolve critical scientific and technological barriers. Using the GOL approach as an alternative to adopting a pre-defined DBH for deciding when to harvest trees in a species-specific manner can greatly improve the silvicultural performance of restoration plantations for timber production. Managing trees based solely on their DBH is ecologically arbitrary since it ignores the need of each species to achieve their maximum productivity for the first harvest and to advance size classes during harvest cycles. Furthermore, the decision to harvest trees based on a fixed DBH could make forest restoration less attractive financially once trees must be maintained despite not growing well. Our results emphasize the need for species-specific growth rate data to optimize the management of restored forests. This ensures a balanced approach between time for harvesting and productivity, leading to reduced harvesting time for similar-sized plants during the first harvest and subsequent management periods. Furthermore, it results in trees of larger diameter within an equivalent growth period.

It is important to recognize, however, that using the GOL approach in restoration, in general, may favor harvesting trees of smaller DBH, which has important consequences for wood processing and quality. Thinner logs has reduced sawing yield and may be composed of a lower proportion of heartwood, which have higher market value due to the greater mechanical stability and resistance to predators and natural hazards (Chave et al., 2009). For example, the teak obtained from plantations in central Brazil has a high proportion of whitish sapwood, as opposed to the dark brown heartwood of larger native teaks exploited from forest remnants in Asia (Carmo et al., 2020; Pereira et al., 2021). Companies have sold whitish teak timber for a lower price but are finding alternative ways to use the timber commercially, considering its particularities (Lima et al., 2021). The regular wood mills that operate using timber exploited from native remnants in the Amazon, as well as those using large logs of eucalypt and pine produced in southern Brazil, may not be adequate to process small native trees coming from restoration plantations. Modern, more sophisticated equipment is required, and novel wood products and uses must be developed to better incorporate the native timber from restoration sites in the market. Recent advances in wood engineering, particularly in processing technology and log utilization rates, hold the potential to significantly broaden the possibilities of using this type of wood for diverse applications. The timber produced in restoration plantations may represent a potential shift from buying timber from unreliable sources, often illegal logging, to a controlled source based on tropical forest restoration.

The superior silvicultural performance of exotic species is a direct consequence of decades of research, development, and innovation performed by public and private agents (Gonçalves et al., 2013). The native tree species investigated here were not submitted to any level of silvicultural improvement and, as expected, still offer a large room for improvement opportunities in forestry sciences. Many valuable timber species historically exploited may not be good candidates for timber production as they grow too slowly. Restoration projects should consider intercropping slow-growing high-quality timber species with species of fast to intermediate growth to offset costs and anticipate revenues (Brancalion et al., 2020; Holl and Aide, 2011). Besides, the production of native timber in restoration plantations will be riskier than using exotic species from industrial monocultures due to the lack of clarity in the business plan (growth rate, payback, future price, and others) and market trend information. For this reason, business-as-usual decisions may not enable the implementation of this system, and more ecological studies about interactions between slow- and fast-growing species from different successional groups are needed to understand species performances and reduce risks. It is critical to consider that restoration plantations will deliver several other benefits, especially those related to biodiversity, soil, and hydrological aspects (Hua et al., 2022), non-timber forest products and other ecosystem services which could also be monetized and "bundled" with timber to offer attractive cash-flow to farmers.

A good perspective exists to achieve sustainable native timber management in forest restoration programs while providing ecosystem services. To achieve high productivity, species-specific criteria related to tree growth, the performance of combined tree species, management, and harvesting plans, as well as silvicultural treatments and research development and innovation must all be incorporated into the forest restoration value chain.

PHSB and RRR are partners of the Re.green company.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. ☒ The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Pedro Medrado Krainovic reports State of Sao Paulo Research Foundation provided financial support. Pedro Medrado Krainovic reports The Nature Conservancy provided financial support. Angélica Faria de Resende reports the State of Sao Paulo Research Foundation provided financial support. Catherine Torres de Almeida reports that the State of Sao Paulo Research Foundation provided financial support. Danilo Roberti Alves de Almeida reports the State of Sao Paulo Research Foundation provided financial support. Pedro Medrado Krainovic reports World Resources Institute provided financial support. Pedro H. S. Brancalion reports The Nature Conservancy provided financial support.