Study shows limits of and proposes alternatives for the functional restoration of the Atlantic Forest

The research used network theory to analyze the ecological connectivity of 28 areas in the northwest of the state of São Paulo, Brazil.

By José Tadeu Arantes | Agência FAPESP – Planting trees is not enough. Although forest restoration efforts in the Atlantic Forest are advancing on a large scale, they are still unable to fully integrate replanted areas into the mosaic of native fragments. This is the conclusion of a study conducted in Brazil and published in the Journal of Applied Ecology.

The first authors of the article were Débora Cristina Rother, a professor at the Federal University of São Carlos (UFSCar) in the state of São Paulo, and Carine Emer, a researcher associated with the Rio de Janeiro Botanical Garden Research Institute and the Juruá Institute in the state of Amazonas. They developed a novel approach based on network theory to analyze the ecological connectivity of 28 areas in the Batatais region in northwestern São Paulo.

According to the researchers, active restoration, or planting seedlings in completely deforested areas, results in plant communities that are separate from the remaining forest fragments. "The restored areas aren't fully integrated into the landscape," Rother summarizes. "What we found was a subset of generalist species connecting the system, mainly small-seeded trees dispersed by birds."

Network theory is a mathematical and computational approach used to understand complex systems formed by many interconnected elements. Rather than analyzing each component in isolation, it represents them as "nodes" and their interactions as "lines," which makes it possible to identify emerging patterns as a whole. This same logic explains the functioning of social networks, biological neural networks (i.e., neural circuits in the nervous system), and artificial neural networks (i.e., computational models inspired by how the brain functions). It can also be applied to ecology in the form of "ecological networks," which analyze the interactions between plants, animals, and environments. This perspective shows not only which species are present in a system but also how they relate to each other and which species are crucial to the system's stability and resilience.

The study used network theory to synthesize a complex database obtained through years of fieldwork. "This is rare data, resulting from a major collective research effort," Emer points out. "Network theory helped us see the whole picture. Instead of focusing on isolated fragments, we looked for interactions between them and restored areas in the landscape."

Structural metrics such as connectance, modularity, and nesting were analyzed. Connectance measures how many connections exist in a network compared to the total possible connections. The higher the connectance, the more species or elements are connected to each other. Modularity indicates the formation of subgroups within the network in which certain elements interact more with each other than with the rest of the system. Nesting occurs when less diverse areas contain subsets of the interactions present in more diverse areas, revealing a hierarchy of inclusion.

"Our networks showed low connectivity values, indicating that few species are widely distributed. Modularity, on the other hand, was intermediate but significant, reflecting the separation between restored areas and native fragments. And nesting, which would indicate whether restored areas could be considered subsets of forests, was very low. This reinforces the idea that restored areas don't yet mirror natural diversity," says Emer.

The study revealed consistent patterns when investigating which species act as “central nodes” in the networks. "We found that key species have two characteristics in common: small seeds and dispersal by animals," Rother explains. "These are plants such as embaúba (Cecropia pachystachya), sangra-d’água (Croton urucurana), tapirira (Tapirira guianensis), and guareia (Guarea guidonia)."

These pioneer trees are essential for initiating natural succession. "They're the first to establish themselves and create conditions for other species to emerge later," says the researcher. "Birds such as thrushes, tanagers, and toucans, as well as small mammals, act as the main dispersers in this process."

Plants such as the embaúba (Cecropia pachystachya) act as "central nodes" in the networks and share two characteristics: small seeds and dispersal by animals (photo: Marcelo Cava/Wikimedia Commons)

Despite advances, restoration faces structural barriers. "We have a huge bottleneck in the production of seedlings from the enormous tropical diversity," says Rother. Important species, such as guareia, are difficult to propagate. "Often, nursery workers try various methods but are unable to germinate the seed."

In addition to technical limitations, there are market barriers. "Nurseries produce what's in demand, and today, restoration is primarily seen as tree planting or carbon capture," Emer notes. "But restoration isn't just about planting trees or storing carbon. We need to restore the ecological processes that ensure the forest functions as a whole, such as the interactions between flora and fauna. If we want to restore diversity, we need subsidies to produce less common species on a large scale and make them available on the market."

Considering these factors, a central point of the study was to demonstrate that planting trees alone is insufficient. While important, it does not solve the problem on its own. "A forest is composed of complex ecological processes," Emer points out. "We also need to look at ecological interactions: birds and mammals dispersing seeds, pollinators ensuring reproduction, succession cycles establishing themselves."

In this sense, techniques such as installing artificial perches to attract birds can help, albeit slowly. Another strategy under discussion is refaunation, or the reintroduction of large dispersers that have disappeared, such as tapirs and agoutis. Rother argues, "If species with large seeds no longer have dispersers, the solution may be to reintroduce them," recalling successful experiences in the Tijuca Forest and the Guapiaçu Ecological Reserve, both of which are located in the state of Rio de Janeiro.

The results have direct implications for public policies and the goals of the United Nations (UN) Decade of Restoration. "The diversity of species used in restoration needs to increase," Emer emphasizes. "We can't limit ourselves to a small set because this compromises the integration of the areas into the landscape as a whole."

Rother believes the strategy should combine pioneer species, which ensure initial succession, with rare species that have large seeds and are unlikely to arrive on their own. "Restoration needs to consider the functional characteristics of plants, not just increase the list of species," she says.

Despite current limitations, the researchers remain optimistic. "The fragments and restored areas form a meta-network," says Emer. "Although it's modular today, this meta-network can become more connected if we increase diversity and encourage interactions."

This study was supported by FAPESP through projects 13/50718-5, 12/24118-8, and 15/15172-7, all of which are part of the Foundation's Research Program on Biodiversity Characterization, Conservation, Restoration, and Sustainable Use (BIOTA-FAPESP).

The article "Plant species-habitat networks in a mosaic landscape of restored and fragmented tropical forests" can be accessed at https://besjournals.onlinelibrary.wiley.com/doi/10.1111/1365-2664.70124.