Resilience of lowland Atlantic forests in a highly fragmented landscape: Insights on the temporal scale of landscape restoration

Highlights

• Tree species richness, composition and carbon stock recover during forest succession.

• Endemic and threatened species richness show no relationship with forest age.

• After 20 yr, secondary forests recovered on average 16–52% of old-growth forests values.

• 20-yr-old secondary forests are potentially sequestering 1.78 Mg C ha⁻¹ yr⁻¹.

• Recovery time predictions are far beyond the targeted by national restoration efforts.

Abstract

The Atlantic forest is one of the most threatened tropical forest ecosystem in the world, and despite current knowledge, its ability to recover structure and diversity after a disturbance is still a matter of debate. Quantifying carbon stocks and species diversity in forests at different successional stages and assessing their recovery capacity is important for designing local conservation and restoration strategies. We investigated the resilience potential of lowland Atlantic forests at landscape-scale using a chronosequence approach in 160 0.1-ha permanent plots, three old-growth and 13 second-growth forests at various stages of recovery. We assessed whether tree species richness, including species with high conservation values (endemic and threatened), composition and aboveground carbon stock recover along succession; and estimated how much of the old-growth forest values the 20-yr-old secondary forests had attained and the time needed to reach old-growth forests levels. Species richness, composition and carbon stock tended to recover along the chronosequence, while endemic and threatened species showed no relationship with forest age. After 20 yr of succession, the secondary forests recovered on average 52% of total species richness, 21% of species composition and only 16% of carbon stock of old-growth forests. We predicted that the absolute recovery of lowland Atlantic forests would take eight decades to thousands of years, much longer than the 10–40 years targeted by national efforts to restore degraded ecosystems. Despite slow recovery, these regenerating forests are important for climate mitigation and biodiversity conservation as they potentially sequester 1.78 Mg C ha⁻¹ yr⁻¹ and harbor a number of species comparable to old-growth forests. Our findings indicate that achieving landscape restoration and conservation goals through passive restoration can be a challenge, highlighting the need to invest in management plans in areas with relatively low resilience and high biodiversity and carbon conservation values.

Introduction

Tropical forests are hotspots for biodiversity and carbon conservation (Beer et al., 2010, Bonan, 2008, Myers et al., 2000). They are subject to several anthropogenic impacts, such as fragmentation, deforestation and selective logging (Malhi et al., 2014), which result in losses of biodiversity (Barlow et al., 2016, Bihn et al., 2008, Gibson et al., 2011, Laurance et al., 2002, Newbold et al., 2015) and large amounts of carbon in the form of the greenhouse gas CO₂ (Baccini et al., 2017, Magnago et al., 2015). For these reasons, deforestation and degradation of tropical forests are especially harmful to the environment and ecosystem functioning (Malhi et al., 2002).

Human impacts have also increased the proportion of second-growth forests across the landscape (FAO, 2018). These secondary forests of different sizes play a crucial role in biodiversity conservation and carbon mitigation, but much debate on these issues and traditional research tend to revolve around mature forests (Gibson et al., 2011, Powers and Marín-Spiotta, 2017, Sullivan et al., 2017). Fortunately, the importance of secondary forests has been increasingly recognized in recent years (Bongers et al., 2015, Chazdon et al., 2016, Poorter et al., 2016), especially those forests able to naturally recover their structure and functions, representing resilient ecosystems (Safar et al., 2019, Willis et al., 2018).

Not all ecosystems are able to recover naturally after a disturbance ceases, because forest recovery and successional trajectory can be influenced by climate (Becknell et al., 2012, Toledo et al., 2011), soil properties (Pinho et al., 2018, Toledo et al., 2018), land use history (Fu et al., 2017, Jakovac et al., 2015) and the initial colonization (Chazdon, 2014). Since trees play a crucial role in the provision of ecosystem functions and services, such as the regulation of global carbon cycle (Bonan, 2008, Oliver et al., 2015), resilient forests ecosystems potentially offer a low-cost approach to climate change mitigation (Busch et al., 2019, Chazdon et al., 2016, Poorter et al., 2016) and biodiversity conservation (Ferreira et al., 2018, Phelps et al., 2012). For these reasons, the conservation of resilient ecosystems has been included in local and international environmental policies and targets inspired by the United Nations Framework Convention on Climate Change (UNFCCC) and the Convention on Biological Diversity (CBD).

There are several global and national efforts to maintain and enhance natural tropical ecosystems within Brazilian landscapes. For example, the Brazilian Atlantic Forest Restoration Pact (AFRP), launched in 2009, aims to promote the restoration of 15 million hectares of the forest by the year 2050 (Melo et al., 2013). Aligned with the Aichi Targets, the Brazilian Biodiversity Targets for 2020, adopted in 2013, aims to enhance ‘ecosystem resilience and the contribution of biodiversity to carbon stocks through conservation and restoration actions, including restoration of at least 15% of degraded ecosystems’ (Aichi Target 15, Strategic Plan for Biodiversity 2011–2020) (CBD, 2013, MMA, 2017). In 2015, the Brazilian government established the National REDD + Strategy (ENREDD +) in order to contribute to climate change mitigation by promoting forest conservation and recovery, and eliminating illegal deforestation by 2020 (MMA, 2016). In the following year, Brazil committed to restore 12 million hectares of land by the year 2030 under the Bonn Challenge, another global forest landscape restoration (FLR) initiative to restore degraded and deforested lands for biodiversity and carbon benefits that represents a practical means of achieving Biodiversity Targets and UNFCCC’s REDD + goals.

Understanding the capacity and the time needed for forests to recover from a disturbance might help to establish safe predictions about what would happen to these forests in different scenarios of environment impacts, and design effective conservation initiatives and investments. Yet, little is known about the recovery rate of the structure and tree composition of the Brazilian Atlantic Forest (see de Paula et al., 2015, Liebsch et al., 2008, Piotto et al., 2009), one of the most diverse and threatened tropical forest in the world (Myers et al., 2000), remaining only 12.4% of forest remnants above 3 ha (SOS Mata Atlântica, 2018). The long history of human occupation and exploitation in the Brazilian Atlantic forest resulted in highly fragmented landscapes, mainly composed of small fragments at some stage of recovery (Ribeiro et al., 2009).

Here we used a space-for-time substitution, or chronosequence approach, to investigate the resilience potential of Brazilian Atlantic forests in a landscape context using data from old-growth and second-growth forests of various ages. We studied the resilience of tree species, including those with high conservation value (endemic and globally threatened species), and the aboveground carbon storage. The definition of resilience adopted here is the capacity of an environment parameter (e.g. species richness) to return to pre-disturbance levels (Nimmo et al., 2015), i.e. the old-growth forests values (reference forests). To accomplish our main goal we (i) investigated whether tree species richness, species composition and carbon stock are recovering along the chronosequence at a landscape scale; (ii) assessed how much of the old-growth forest values the 20-yr-old secondary forests had attained, and (iii) estimated the absolute and relative time that second-growth forests take to approach the tree composition and mean old-growth forests values, used as a proxy of full recovery (100%). Then, we discuss the implications of our findings for biodiversity and carbon conservation and mitigation in human-modified landscapes under global and local initiatives for ecological restoration.