Changes like the shift of tropical forests into savannah in the Amazon highlight the potential for deforestation to drive ecosystems past potentially irreversible tipping points. Reforestation may avert or delay tipping points, but its success depends on the degree to which secondary and primary forests are substitutes in the production of ecosystem services. This article explores how deforestation, reforestation and substitutability between forest types affect the likelihood that a forest system will cross a tipping point. Efforts to ensure that secondary forests better mimic primary forests only yield a small improvement in terms of delaying ecosystem collapse. The most significant effects on tipping points arise from an increase in the relative costs of clearing primary forests or a decrease in the costs of protecting land tenure in secondary forests. Our results highlight the importance of the latter, which are often ignored as a policy target, to reduce the risk of ecosystem collapse.

The endemic Crow Honeyeater Eugymnomyza aubryana of New Caledonia is classified as “Critically Endangered”. Its rainforest habitat on mainland Grande Terre has been reduced by around 80% and the remaining habitat is highly fragmented. It is likely that by early 1900, anthropogenic changes had split the Crow Honeyeater’s distribution into two populations: one in the north of the island and one in the south separated by over 200 km. From 2011, the species was only known to be present in the south of Grande Terre with its distribution centred on the Blue River Provincial Park. It is essential for the species’ survival to protect its remaining natural habitat and its nests from predators.

Phalangerids in Sulawesi occur at the westernmost extent of marsupial distribution in Wallacea and are facing escalating anthropogenic pressures. The ecology of the Vulnerable bear cuscus Ailurops ursinus, one of four endemic phalangerids in Sulawesi, is poorly understood and the extent of its decline unknown. This study aimed to build the first habitat suitability model for the species in South Sulawesi and identify priority areas for its conservation. We used maximum entropy modelling to predict the species’ potential distribution, and overlaid the resulting habitat suitability map with regulatory and mining activity maps. Our model predicted only 7.5% (143,682 ha) of the total study area to be potentially suitable habitat for the bear cuscus, predominantly in scattered forest patches, some of which are in areas affected by active mining or frequent poaching. Land-cover type was the most important predictor of the species’ distribution. Our findings suggest the lack of legal protection for the bear cuscus should be reconsidered, and we recommend the species is reassessed for the IUCN Red List. The forested areas of southern South Sulawesi have decreased by 12.5% since 2000, with over half of this decline occurring since 2015, preventing northwards dispersal and restricting the species to a fraction of the province's protected forests. Immediate intervention is necessary to combat poaching, slow the expansion of mining and increase landscape connectivity, to prevent further reduction of the species’ current and potential habitat.

The Northeast Japan triple disasters have raised questions about the adequacy of government planning for water damage at Fukushima, yet other components of the built environment also contributed to loss of property and life in unanticipated ways. Such “surprises” associated with large-scale construction have a long history. To illustrate, in addition to exploring ways in which the built environment shaped the tsunami's impact, I examine negative consequences from the construction of the Okotsu Diversion Channel (Niigata). Through the example of a major flood in Tochio (Niigata), I show unexpected links to a government policy, the Meiji Land Tax Reforms, that did not deal with water control at all.

Cocoa is an important agricultural product that plays a crucial role in local communities in South America. In Brazil, it is traditionally grown in agroforestry systems, which are more sustainable and contribute to biodiversity conservation. However, the recent expansion of intensive monocultures in tropical forests poses significant threats to this activity. Using historical data on land use and cocoa productivity at the municipality level from Brazil’s primary cocoa-producing states, we show that maintaining and restoring forest cover are positively correlated with cocoa productivity, particularly in areas with less anthropogenic disturbance. This highlights the dependence of cocoa production on ecosystem services. Recent data reveal that in municipalities where local agriculture is less reliant on cocoa, only larger farms have benefitted from increased forest cover, probably due to their greater dependence on fragments of natural habitat for ecosystem services. In municipalities that are more reliant on cocoa, the effects of forest cover were not detected, while strong negative effects of forest fragmentation were observed in both small- and large-scale farms. We emphasize the importance of preserving natural forests near cocoa plantations to optimize productivity in Amazon and Atlantic Forest agroforestry, especially in deforested areas.

The additionality of forest conservation interventions is frequently questioned. In particular, they are often considered to be located in places where forests are not threatened, which points to the existence of location biases. Revisiting this location bias concept, we conceptually distinguish potential and effective additionality and theoretically consider how the objectives of the implementer affect the siting choice of the forest conservation interventions and their additionality. Our theoretical intuition is that the choices of the implementers are influenced by the quality of institutions. Our results show that (1) the implementer's objective and local institutions may lead the implementer to select a site with low development potential and low forest threat, and (2) the selection of a site with low development potential, which is frequently presented as a location bias, does not necessarily preclude additionality.

In West Africa, vast areas are being deforested; the remnant forest patches provide a wealth of ecosystem services and biodiversity conservation potential, yet they are threatened by human activity. Forest patches <100 ha have not been widely catalogued before; we mapped forest loss of small forest patches outside of protected areas in the Guinean savannah and humid Guineo-Congolian bioclimatic regions of Togo, Benin, Nigeria and Cameroon between 2000 and 2022. Focusing on the dynamics of small patches, without considering the splitting process of larger patches, we quantified changes in their number and area and the rate and trend of forest loss. Small forest patches are widespread, yet their area and number have decreased, while the forest loss rate is increasing. Primary forest patches lost almost half of their area annually – twice as much as secondary forests, and this loss was especially pronounced across small patches (0.5 – 10 ha), suggesting deforestation preferentially occurs in the smallest patches of primary forest. If forest loss continues at the current rate, 14% of the total forest area mapped in this study will have disappeared by 2032, jeopardizing their potential to provide ecosystem services and emphasizing the need for measures to counter their deforestation.

Conflicts over resources with poorly defined property rights have fuelled both deforestation and violence in the Brazilian Amazon. However, what happens when the State enhances its ability to monitor and enforce existing environmental laws? We study the case of the list of Municípios Prioritários, a policy that allocates additional resources to verify compliance with environmental laws in municipalities with high deforestation rates. Employing a difference-in-differences approach, our findings suggest that an improvement in the ability of the State to monitor and enforce environmental laws can reduce conflicts over the appropriation of value from resources with poorly defined property rights. Consistent with existing studies, we also find that the policy led to a reduction in deforestation rates in the Brazilian Amazon. Finally, we discuss the limitations of the current approach to curb violence in a region where the activity of mafias has considerably grown since the turn of the twenty-first century.

The amphibians of the Indonesian island of Sumatra are poorly known, despite it being recognized as a biodiversity hotspot. For determining conservation priorities, up-to-date knowledge of the state of amphibian diversity in Sumatra is crucial, particularly considering the high deforestation rate on the island. To address this, I compiled and analysed a comprehensive dataset for amphibians known from Sumatra, to identify knowledge gaps and assess the significance of these data for conservation on the island and, more broadly, across Indonesia. The compilation indicates there are 135 amphibian species currently recorded for Sumatra, 55% more than the number known in 1923. Approximately 44 species have been described since 2000. Associated data on larvae and calls is lacking for many species. Although 66% of the amphibians of Sumatra are categorized as Least Concern on the IUCN Red List, this may not represent the actual conservation status of these species because many of the assessments are outdated. In addition, 14% of species have not yet been evaluated. This study highlights the need for more comprehensive studies on the amphibians of Sumatra.

The concept of a forest transition – a regional shift from deforestation to forest recovery – tends to equate forest area expansion with sustainability, assuming that more forest is good for people and the environment. To promote debate and more just and ecologically sustainable outcomes during this period of intense focus on forests (such as the United Nations’ Decade on Ecological Restoration, the Trillion Trees initiative and at the United Nations’ Climate Change Conferences), we synthesize recent nuanced and integrated research to inform forest management and restoration in the future. Our results reveal nine pitfalls to assuming forest transitions and sustainability are automatically linked. The pitfalls are as follows: (1) fixating on forest quantity instead of quality; (2) masking local diversity with large-scale trends; (3) expecting U-shaped temporal trends of forest change; (4) failing to account for irreversibility; (5) framing categories and concepts as universal/neutral; (6) diverting attention from the simplification of forestlands into single-purpose conservation forests or intensive production lands; (7) neglecting social power transitions and dispossessions; (8) neglecting productivism as the hidden driving force; and (9) ignoring local agency and sentiments. We develop and illustrate these pitfalls with local- and national-level evidence from Southeast Asia and outline forward-looking recommendations for research and policy to address them. Forest transition research that neglects these pitfalls risks legitimizing unsustainable and unjust policies and programmes of forest restoration or tree planting.

We examine how agricultural subsidies may induce deforestation and interact with conservation programs by analyzing two large-scale national programs in Mexico that have existed simultaneously for more than a decade: an agricultural subsidy for livestock (PROGAN) and a program of payments for ecosystem services (PES). Looking across the entire Mexican landscape, we exploit the surprises in the timing of enrollment in PROGAN's waves, fluctuations in program payments, and the change in the value of the subsidy induced by inflation and currency fluctuations to identify the impacts of the livestock subsidy on environmental outcomes. We find that PROGAN increased municipal deforestation by 7 per cent. The deforestation effects of PROGAN were smaller in municipalities with higher concentrations of PES recipients. We suggest that livestock subsidies could be better targeted to places with low deforestation risk and high livestock productivity to maximize food production and minimize negative externalities caused by deforestation.

The Upper Guinea Forest (UGF; West Africa), a global biodiversity hotspot, has lost more than 90% of its original area since 1900, threatening endemic species such as the endangered pygmy hippopotamus (Choeropsis liberiensis). However, little is known about the proximate causes of this deforestation. We classified Sentinel-2 data using the random forest algorithm to differentiate between three main human processes (shifting agriculture, intensive agriculture or urban expansion) driving deforestation between 2000 and 2019 across the pygmy hippopotamus distribution area. Out of c. 89 600 km2 in the year 2000, 15 900 km2 (17%) of forest were lost, primarily to shifting agriculture (14 900 km2). Côte d’Ivoire and Liberia accounted for 14 900 km2 (94%) of the net area of forest lost, c. 15 times greater than deforestation in Sierra Leone and Guinea combined (953 km2). Forest loss inside protected areas is pervasive, and it is essential to prioritize conservation efforts in areas where deforestation is still low (e.g., Taï, Sapo and Gola Rainforest national parks). We suggest that the preservation of the UGF will face challenges associated with people’s demand for food and income. Continued landscape-scale planning and action to reduce deforestation are urgently needed to limit the impact of shifting agriculture on pygmy hippopotamus habitat.

The COVID-19 pandemic and government responses led to a halt in economic activity. While this reduced pollution in urban areas, its effect on deforestation in areas outside of cities is unclear. Deforestation may have decreased due to the restrictions on economic activity, but, it may have increased due to the drying up of alternative income sources. We analyzed bi-weekly data on tropical forests worldwide in relation to the dates when different countries implemented lockdown restrictions. Our analysis found that while lockdowns did reduce mobility in forest municipalities, the average effect on deforestation was not significant. However, we did observe variations in the impact of lockdowns on deforestation based on the share of lockdown-vulnerable GDP and the level of government effectiveness. These results stand across tropical countries and within Colombia. These findings highlight the importance of alternative income sources and strong state capacity for effective policies aimed at reducing deforestation.

Predicting future conservation needs can help inform conservation management but is subject to uncertainty. We measured deforestation rates during 2015–2017 for 114 protected areas in Madagascar, linked deforestation to the status of protection according to IUCN categories I–VI, used recent deforestation rates to extrapolate forest cover over 2017–2050 and linked the size of forest blocks to the projected persistence of lemur subpopulations. In the six IUCN categories for protected areas in Madagascar the median size of forest blocks is 9–37 km2 and median annual deforestation rates range from 0.02% in the single IUCN category III site to 0.19% in category II and 1.95% in category VI sites. In 2017, 40% of all forest blocks within protected areas were < 10 km2, and this is projected to increase to 45% in 2050. Apart from these small forest fragments, the modal site of forest blocks was 160–320 km2 in 2017, and this is projected to decrease to 80–160 km2 in 2050. The range of > 50% of all lemur species exclusively contains forest blocks of < 10 km2. The modal size of forest blocks > 10 km2 is predicted to remain at 120 km2 until 2050. Although uncertainty remains, these analyses provide hope that forest blocks within the protected areas of Madagascar will remain large enough to maintain lemur subpopulations for most species until 2050. This should allow sufficient time for the implementation of effective conservation measures.

This article examines whether mass deforestation could be prosecuted as a crime against humanity under Article 7 of the Rome Statute. It does so in respect of the situation in the Brazilian Legal Amazon in 2019–2021, where the unbridled exploitation and destruction of the rainforest had a disastrous impact at local, regional and global levels. The article covers three main aspects. First, it explores the existing limits of international criminal law for prosecuting mass deforestation as a crime against humanity, and the contours within which criminalization would be possible. Secondly, it discusses the challenges inherent in the anthropocentric nature of the chapeau requirement of Article 7 for the criminalization of mass deforestation under that provision. Thirdly, it analyses the extent to which mass deforestation could qualify as persecution and/or an ‘other inhumane act’ under Articles 7(1)(h) and (k) of the Rome Statute.