Editing IPCC:AR6/WGII/Cross-Chapter-Paper-7 (section)

== CCP7.3 Current and Projected Climate Change Impacts on Tropical Forests (Drought, Temperature, Extreme Events) ==

<div id="h1-4-siblings" class="h1-siblings"></div>

While early dynamic global vegetation models predicted biome shifts and contractions of tropical forests, more recent efforts have focused on biome changes at more regional scales, or on functional aspects of tropical forests, such as plant physiological and phenological changes, drought-related mortality, population dynamics, interspecies interactions and community responses, ecohydrology, risk of fire and related impacts, soil nutrient and microbe–plant interactions. Climate change is expected to increase temperatures across the tropics, with attendant variability in rainfall, and more extreme events such as intense storms, droughts and wildfires (Zelazowski et al., 2011; Malhi et al., 2014; Brando et al., 2019). This could be expected to have structural and functional impacts on tropical forest biomes (Malhi et al., 2014; Adams et al., 2017). This section looks at responses of tropical trees and forests to current and future climate-change related pressures, focusing on physiological responses including growth, mortality and regeneration, fire risk and ecological vulnerability, as well as on climate effects of tropical forest loss.

<div id="CCP7.3.1" class="h2-container"></div>

<span id="ccp7.3.1-tropical-tree-physiological-responses-to-climate-change"></span>
=== CCP7.3.1 Tropical Tree Physiological Responses to Climate Change ===

<div id="h2-4-siblings" class="h2-siblings"></div>

With rising temperatures and atmospheric carbon dioxide, possibly accompanied by greater variability in soil moisture availability, a key question is how tropical forest trees respond physiologically (especially photosynthesis and respiration which determine net growth rates) and how well they can acclimate (i.e., able to adapt) to climate change (Dusenge et al., 2019). Key climate factors influencing tree growth on pan-tropical forests are precipitation, solar radiation, temperature amplitude and relative soil moisture (Wagner et al., 2014).

The temperature response of photosynthetic carbon uptake in tropical trees seems remarkably similar across moist and dry forest types, as well as for light-demanding, fast-growing species compared with shade-tolerant, slow-growing species ( [[#Slot--2017|Slot and Winter, 2017]] ). It is generally agreed that photosynthesis in tropical species can acclimate to moderate levels of warming but beyond this there would be no net gain in carbon ( [[#Slot--2017|Slot and Winter, 2017]] ). The factor that limits photosynthesis in different tropical forests will depend on water availability. In water-limited dry forests, photosynthesis may decline largely due to stomatal closure, while in wet forests the decline may largely be driven by warming-related changes to leaf biochemistry ( [[#Slot--2017|Slot and Winter, 2017]] ). A recent modelling approach suggests that the limits of photosynthetic thermal acclimation may be an increase of about 2°C, in terms of maximum tolerated temperature, with enhanced tree mortality beyond this level of warming ( [[#Sterck--2016|Sterck et al., 2016]] ).

A critical concern for plant function has been that higher temperatures will enhance respiration rates, potentially resulting in tropical forests becoming net carbon sources (rather than photosynthesis-driven carbon sinks) ( [[#Gatti--2021|Gatti et al., 2021]] ). Some studies suggest that excessive respiration is less of a concern as respiration rates can acclimate to elevated temperatures over time ( [[#Lombardozzi--2015|Lombardozzi et al., 2015]] ; [[#Pau--2018|Pau et al., 2018]] ). Thermal acclimation of respiration has been shown in a seasonally dry neotropical forest ( [[#Slot--2014|Slot et al., 2014]] ), while models indicate that increases in plant respiration could halve by the end of the 21st century through acclimation, thereby partly ameliorating the potential release of carbon from tropical forests ( [[#Vanderwel--2015|Vanderwel et al., 2015]] ). A contrary view is that plant physiological processes, such as the photosynthesis in tropical canopy trees, are already functioning at levels close to or beyond their thermal optimum limits and that any further temperature increase would turn them from a sink into a carbon source ( [[#Mau--2018|Mau et al., 2018]] ). One of the most pressing questions regarding forest responses to increasing atmospheric CO 2 levels is whether trees experience enhanced growth rates as a result of the so-called CO 2 fertilisation effect [Box 2.3 in IPCC 2019b]. Observed changes in the terrestrial carbon sink and process-based vegetation models indicate that tropical vegetation response to CO 2 fertilisation ( [[#Schimel--2015|Schimel et al., 2015]] ) is combined with other factors such as nitrogen deposition and length of the growing season, while aerosol-induced cooling may also have played a role in enhancing the carbon sink [Box 2.3 in IPCC 2019b]. Contrastingly, evidence for CO 2 fertilisation of growth in individual tropical tree species is generally lacking or controversial ( [[#Silva--2013|Silva and Anand, 2013]] ), or not as substantial as expected ( [[#Sampaio--2021|Sampaio et al., 2021]] ). It is, however, widely agreed that the intrinsic water-use efficiency of a tree, that is, the amount of carbon assimilated as biomass per unit of water used, increases under elevated atmospheric CO 2 levels owing to the regulation of stomata (cells on the leaf surface which regulate the exchange of water and gases between the plant and the atmosphere) ( [[#Van%20Der%20Sleen--2015|Van Der Sleen et al., 2015]] ; [[#Bartlett--2016|Bartlett et al., 2016]] ; [[#Rahman--2016|Rahman and Alam, 2016]] ; [[#Keeling--2017|Keeling et al., 2017]] ). Tropical dry forests (ca. 1000 mm annual rainfall) exhibit changes in water-use efficiency (WUE), relative to CO 2 , at least twice as much as tropical moist forests (c. 4000 mm rainfall) ( [[#Adams--2019|Adams et al., 2019]] ).

Other key components in the forest system are plant–microbe–soil nutrient interactions, which play major roles in carbon cycling and plant photosynthetic response to increased atmospheric CO 2 and warming (Zhang et al., 2014; [[#Singh--2015|Singh and Singh, 2015]] ; Du et al., 2019). Phosphorus is generally a limiting factor in tropical forest soils, though this may be species-specific (Ellsworth et al., 2017; Turner et al., 2018). Mycorrhizal fungi (both arbuscular and ectomycorrhizal) play major roles in water acquisition of host plant and their responses to drought in dry tropical forest ( [[#Lehto--2011|Lehto and Zwiazek, 2011]] ) as well as in the capture and transfer of nutrients, especially nitrogen (which may otherwise become limiting), to host plants. Climate change factors can thus be expected to alter the nature of soil–plant interactions with consequences for the species composition and biodiversity of tropical ecosystems (Pugnaire et al., 2019; Terrer et al., 2019)

<div id="CCP7.3.2" class="h2-container"></div>

<span id="ccp7.3.2-climate-related-mortality-and-regeneration-in-tropical-forests"></span>
=== CCP7.3.2 Climate-Related Mortality and Regeneration in Tropical Forests ===

<div id="h2-5-siblings" class="h2-siblings"></div>

Drought-related mortality of tropical trees shows complex patterns which could change forest community structure and composition with cascading effects on biodiversity (McDowell et al., 2020). During drought, the mortality rate is enhanced in larger-sized trees in tropical forests (as is the case with all forests globally), with significant impacts on forest structure, carbon storage and regional hydrology (Bennett et al., 2015). The mortality rate of neotropical moist forest trees appears to be consistently increasing since the 1980s (McDowell et al., 2020), with plant functional types such as softwood, pioneer and evergreen species suffering higher mortality during years of extreme drought (Aleixo et al., 2019). Large trees (&gt;30 cm diameter at breast height (dbh)) in tropical dry forests have much lower mortality rates than those reported for tropical moist forests (Suresh et al., 2010). Contrary to expectation, during prolonged droughts in these dry forests, deeper-rooted tree species are more ''likely'' to die than shallow-rooted ones, which are more adapted to changes in soil moisture content, because of water depletion in the deepest unsaturated zone (Chitra-Tarak et al., 2018).

Regeneration of tropical tree seedlings and their response to a changing climate is inadequately understood. Experimental work suggests that tropical moist forest tree seedlings and saplings can acclimate photosynthetically to moderate levels of warming and, unlike adults, may even exhibit increased growth rates ( [[#Cheesman--2013|Cheesman and Winter, 2013]] ; [[#Slot--2018|Slot and Winter, 2018]] ). Some moist forest seedlings also show plasticity to recurrent drought episodes by enhancing their growth rates when favourable moisture conditions return, while others fail to respond (O’Brien et al., 2017). The nature of response also seems to be mediated by neighbourhood diversity, with greater plasticity in more diverse communities (O’Brien et al., 2017). Seedlings in tropical dry forests subject to burning show enhanced growth rates post-fire and within two years attain similar height of seedlings in unburnt areas (Pulla et al., 2015), though the environmental drivers of seedling growth post-fire are not well understood (Bhadouria et al., 2017).

The net outcome of the population dynamics processes of growth, mortality and regeneration is change in species composition as a consequence of a changing climate. In the Amazon forests, dry habitat-affiliated genera have become more abundant among the newly recruited trees, while the mortality of moist habitat-affiliated genera has increased in places where the dry season has intensified most, thus driving a slow shift towards a drier forest type (Esquivel-Muelbert et al., 2019). A similar multi-decadal shift in West-African forest species composition towards more dry-affiliated species as a response to long-term drying has been recorded (Aguirre-Gutiérrez et al., 2020). While upward shifts in the tree line and in the range of individual tree species have been recorded at several temperate mountain regions, evidence from the tropics is rare. A large-scale study from 200 plot inventories of &gt;2000 tree species across a ~3000 m elevation gradient in the Andean tropics and sub-tropics has shown that the relative abundances of tree species from lower, warmer locations were increasing at these sites indicating that ‘thermophilisation of vegetation’ (increased domination of plant species from warmer locations) was indeed taking place as expected (Fadrique et al., 2018) [Section 2.5.4.2.1 in Chapter 2].

<div id="_idContainer035" class="Figure"></div>

[[File:c62eca070141c77f9b30de40b14d9f42 IPCC_AR6_WGII_Figure_CCP7_004.png]]

'''Figure CCP7.4 |''' '''Documented instances of tree mortality in tropical moist forests due to fire (1992–2016) and drought (1982–2005).''' These occurrences were associated with anomalies in precipitation and temperature over the study period. Adapted from Brando et al. (2019).

<div id="CCP7.3.3" class="h2-container"></div>

<span id="ccp7.3.3-fire-risks-from-climate-change-in-tropical-forests"></span>
=== CCP7.3.3 Fire Risks from Climate Change in Tropical Forests ===

<div id="h2-6-siblings" class="h2-siblings"></div>

Temperature rise and prolonged droughts increase the danger of fires in drained peatlands and tropical forests in Southeast Asia and the Amazon (da Silva et al., 2018; Pan et al., 2018; Sullivan Martin et al., 2020), resulting in large carbon emissions, which reached 11.3 Tg CO 2 day −1 during September–October 2015 (Huijnen et al., 2016; Yin et al., 2020) and changes in forest composition and biodiversity (Asner et al., 2000; Hoffmann et al., 2003) ( ''high confidence'' ). In many cases, tree mortality due to fire is poorly recorded in the literature, but the available data suggest that fire-induced mortality has increased in recent years (Figure CCP7.2) (Malhi et al., 2014; Brando et al., 2019) ( ''high confidence'' ). While large forest and peat fires used to be associated mainly with El Niño-Southern Oscillation (ENSO) events, there is now evidence that tropical rainforests in Indonesia may experience higher fire danger from increased temperatures even during non-drought years due to high evaporation rates of fragmented forests (Fernandes et al., 2017; McAlpine et al., 2018). The droughts of 2007 and 2010 in the Amazonian region caused 12% and 5% of the southeastern Amazon forests to burn, respectively, as compared with &lt;1% of these forests burning during non-drought years (Brando et al., 2014; da Silva Júnior et al., 2019; Pontes-Lopes et al., 2021). Moreover, degraded forests in Ghana are more vulnerable to fires during droughts (Dwomoh et al., 2019).

Factors other than solely climate also interact in enhancing the danger of tropical forest fires. For instance, the extent of burned area of rainforests in Borneo has shown that subsurface hydrology, (i.e., hydrological drought), interacts with meteorological drought and, hence, fires have become more intense in recent decades following the progressive desiccation of the island over the past century (Taufik et al., 2017). Bornean forest fire risk also increased through the interaction of drought with land use conversion for logging, oil palm and tree plantations, and human settlements (Sloan et al., 2017). Similarly, simulations of future fire risks in the Amazon show that extensive land use change under the RCP 8.5 scenario results in 4- to 28-fold enhanced area of forest burned by fire by 2080–2100, as compared with 1990–2010, whereas in an RCP 4.5 scenario, the area burned would be enhanced by 0.9- to 5.4-fold (Le Page et al., 2017).

<div id="CCP7.3.4" class="h2-container"></div>

<span id="ccp7.3.4-current-climate-risks-for-tropical-forests"></span>
=== CCP7.3.4 Current Climate Risks for Tropical Forests ===

<div id="h2-7-siblings" class="h2-siblings"></div>

Impacts of climate change on tropical forest cover seem to correlate with climatic zone. Natural selection of drought tolerant species is observed in tropical dry forests under a prolonged water deficit environment ( [[#Stan--2019|Stan and Sanchez-Azofeifa, 2019]] ). Tropical montane forests are highly sensitive to warming and associated changes in cloud cover and moisture, with evidence that such forests are already being impacted through ‘browning’ (loss of biomass) from increased warming since the 1990s (Krishnaswamy et al., 2014).

Besides higher temperatures, current climate risks also depend on regional responses to a variety of climate events. For example, tropical biomes across the three continents may respond differently to ENSO events in terms of carbon fluxes and balance. During the 2015–2016 ENSO event, different processes were dominant for the carbon fluxes anomaly in the tropical regions. In Asian forests, this anomaly was primarily derived from enhanced fire occurrence, in African forests through increased ecosystem respiration (from higher temperatures), and in South American forests by ecophysiological effects, through the gross primary production (GPP) expressed as reduced carbon uptake (Liu et al., 2017; van Schaik et al., 2018). It has also been shown that the probability of drought spells at the beginning and end of the rainy season is higher in the areas with the highest deforestation (Leite-Filho et al., 2019). Furthermore, it has been observed that Amazon rainforest resilience is being lost faster in regions with less rainfall and in parts of the rainforest that are closer to human activity ( [[#IPCC--2014|IPCC, 2014]] ; Seiler et al., 2015) (CCP7.3.6). Conversely, it has been pointed out, on the basis of vegetation indices, that temperature has a greater influence on resilience than does precipitation, and tropical forests are more resilient to climate change when they are more diverse (Feng et al., 2021) (CCP7.3.6).

Biomes such as seasonally dry tropical forests subject to higher variability in rainfall or other climatic factors may be more resilient to fire and drought (Pulla et al., 2015; Liu et al., 2017), though there could be changes in species distributions as a result of disturbances (Allen et al., 2017). A regime of long-term, high rainfall variability seems to be critical in determining the overall resilience of tropical forests and savannas to climate disturbances (Ciemer et al., 2019), highlighting the heterogeneity of the tropical landscape to climate risk. Similarly, forest composition, nutrient limitations and biodiversity can influence forest resilience to disturbances. Recent evidence suggests that the degree of forest disturbance also affects the mechanisms through which biodiversity influences forest functioning (Schmitt et al., 2020). Neotropical secondary forests also showed high resilience by maintaining their biomass through high productivity and rates of recovery following major disturbances (Poorter et al., 2016). However, the possibility of tropical forests reaching ‘tipping points’ in their resilience and experiencing rapid die-off cannot be ruled out (Verbesselt et al., 2016).

<div id="CCP7.3.5" class="h2-container"></div>

<span id="ccp7.3.5-projected-impacts-of-climate-change-on-tropical-forest"></span>
=== CCP7.3.5 Projected Impacts of Climate Change on Tropical Forest ===

<div id="h2-8-siblings" class="h2-siblings"></div>

Climate change projections indicate increased warming and changes in rainfall patterns in the tropical region as elsewhere globally (IPCC, 2021, AR6 WGI). These would have impacts on carbon stocks ( [[#Mitchard--2018|Mitchard, 2018]] ; Hubau et al., 2020), water availability (Tamoffo et al., 2019), and structure and diversity (Malhi et al., 2014; McDowell et al., 2020) in tropical forests, amplified by deforestation (CCP7.3.6).

Tropical forests are critical repositories of global carbon; living tropical trees are estimated to hold 200–300 Pg C or about one-third of the levels in the atmosphere ( [[#Mitchard--2018|Mitchard, 2018]] ). CMIP5 and CMIP6 Earth System Models (ESM) project an increasing future tropical carbon sink, which is particularly strong in the scenarios with more pronounced increases in atmospheric CO 2 concentration (Koch et al., 2021). However, major uncertainties regarding the ecophysiological processes governing carbon turnover and tree mortality under a changing climate ( [[#Hartmann--2015|Hartmann et al., 2015]] ; [[#Pugh--2020|Pugh et al., 2020]] ), and the ecosystem-level responses of tropical forests to elevated atmospheric CO 2 ( [[#Körner--2009|Körner, 2009]] ) explain the contrast between observational data and modelling results ( [[#Rammig--2021|Rammig and Lapola, 2021]] ). Observational data show that structurally intact old-growth tropical forests have been net sinks of atmospheric carbon in recent decades, but there is evidence that the capacity of such intact tropical forests to build up carbon stock may be limited as biomass peaked during the 1990s and has since weakened by 30% in the Amazon since the 1990s ( ''high confidence'' ), mainly due to increased tree mortality and faster carbon turnover, and the African tropical forest sink following this trend since about 2010 ( [[#Hubau--2020|Hubau et al., 2020]] ; [[#Gatti--2021|Gatti et al., 2021]] ). From a peak pan-tropical (Amazonia, Africa and Southeast Asia) forest sink of 1.26 Pg C yr −1 during the 1990s, it is projected to decline to an uptake of only 0.29 Pg C yr −1 , reaching zero in the Amazon, during the 2030s ( [[#Hubau--2020|Hubau et al., 2020]] ). This decline will possibly be driven by the reduced rates of forest carbon uptake from the weakening global CO 2 fertilisation effect mediated by limiting soil nutrient, and reduced water availability and higher temperatures during extreme droughts ( [[#Qie--2017|Qie et al., 2017]] ; [[#Fleischer--2019|Fleischer et al., 2019]] ; [[#Wang--2020|Wang et al., 2020]] ), reinforced by deforestation and forest degradation [IPCC SRCCL, 2019].

Offline (uncoupled) vegetation model simulations indicate that the extensive tropical and subtropical forests of the Americas could gradually transit towards a savanna-like vegetation, with the most pronounced shifts (of up to 600 km northward) from relatively stable forests to savanna-forest transitions occurring in the eastern Amazonian region (Huntingford et al., 2013; Anadon et al., 2014; Nobre et al., 2016) depending largely on the yet uncertain strength of the CO 2 fertilisation effect and future dry season length, with important feedbacks on the flux of moisture from the forest to the atmosphere (Zemp et al., 2017). More limited simulations for Central American rainforests under RCP 4.5 and 8.5 also support a transition in some areas to lower biomass tropical dry forest and savanna-like vegetation (Lyra et al., 2017). Such transitions from one biome type to another will cause major changes in forest structure, species compositions and overall biodiversity. Additionally, the difficulty of species to migrate through highly fragmented tropical forested regions (such as West Africa or South and Southeast Asia) and ‘non-analogue climates’, under a climate change scenario, poses extra pressure on tropical biodiversity to adapt and survive (Pörtner et al., 2021). Even in expansive tracts of forests, such as in the Amazon, climate change is expected to become more important than deforestation by 2050 in causing the loss of tree species (Gomes et al., 2019). Tropical mountain biodiversity hotspots (e.g., Andes, Himalayas) are particularly vulnerable to species loss due to elevation range shifts (Sekercioglu et al., 2008). Under a 2°C increase scenario, a substantial reduction of tropical montane cloud forest in Kenya is estimated (Los et al., 2019).

<div id="CCP7.3.6" class="h2-container"></div>

<span id="ccp7.3.6-climate-responses-to-tropical-deforestation-and-links-to-forest-resilience"></span>
=== CCP7.3.6 Climate Responses to Tropical Deforestation and Links to Forest Resilience ===

<div id="h2-9-siblings" class="h2-siblings"></div>

Since AR5, there has been meaningful advancement in understanding the climate effects of deforestation and concomitant changes in forest ecosystem resilience. The IPCC Special Report on Climate Change and Land (Jia et al., 2019) and IPCC AR6 WGI (Douville et al., 2021) both describe significant climate-related changes resulting from tropical deforestation ( ''high confidence'' ).

Deforestation generally reduces rainfall and enhances temperatures and landscape dryness; effects that increase with the scale of forest loss, whereas reforestation and afforestation generally reverses these effects ( ''high confidence'' ) ( [[#Lawrence--2015|Lawrence and Vandecar, 2015]] ; [[#Alkama--2016|Alkama and Cescatti, 2016]] ; Khanna et al., 2017; Jia et al., 2019; Staal et al., 2020; Douville et al., 2021; Hofmann et al., 2021; Leite-Filho et al., 2021). There is also ''medium evidence'' from observations and modelling that deforestation enhances surface runoff (Douville et al., 2021). Whereas quantitative information is much more limited for other tropical regions, past deforestation in the Amazon has led to a small reduction in rainfall of −2.3% to −1.3%, shortening and delay of the wet season, and an estimated 4% increase in dryness (Leite-Filho et al., 2020; Staal et al., 2020; Douville et al., 2021).

Modelling studies estimate that large-scale tropical deforestation will contribute to average warming of the deforested areas with +0.61 ± 0.48°C and will lead to large changes in diurnal temperature ranges owing to a reduction of nocturnal cooling ( ''medium confidence'' ) (Jia et al., 2019). Large-scale deforestation will also strongly decrease average regional precipitation and evapotranspiration and further delay the onset of the wet season, enhancing the chance of dry spells and intensifying dry seasons, but the magnitude of the decline depends on the scale and type of land-cover change ( ''high confidence'' ) (Zemp et al., 2017; Jia et al., 2019; Douville et al., 2021; Gatti et al., 2021).

Continued forest landscape drying and fragmentation in connection with deforestation may also enhance surface flow variability (Farinosi et al., 2019; Souza et al., 2019) and will aggravate the risk of forest dieback (Zemp et al., 2017), elevate forest flammability (Alencar et al., 2015) and increase fire incidence ( ''high confidence'' ) (Aragão et al., 2018; Jia et al., 2019; Silveira et al., 2020; dos Reis et al., 2021), ultimately leading to savannisation of many tropical rainforests (Sales et al., 2020). However, compositional heterogeneity and diversity of forest assemblages increases resilience against climate-enhanced forest degradation (Réjou-Méchain et al., 2021).

For the Amazon, deforestation (ca. 40% of the region) in combination with climate change will raise the prospect of passing a tipping point leading to large-scale savannisation of the rainforest biome, but uncertainty remains whether this will take place in the 21st century (Nobre et al., 2016; Jia et al., 2019; Douville et al., 2021). However, considering that the Amazon has already lost ca. 20% of its forests (Nobre et al., 2016), crossing the tipping point may not only create savannas of the deforested parts but may also result in precipitation reductions of 40% in non-deforested parts of the western Amazon due to a breakdown of the South American monsoonal circulation and the subsequent western cascade of precipitation and evapotranspiration (Boers et al., 2017). Other effects of forest degradation include loss of ecosystem services, biodiversity, carbon storage and Indigenous culture (Watson et al., 2018; Strassburg et al., 2019; Gatti et al., 2021), as well as potentially reduced hydropower capacity and agricultural production (Sumila et al., 2017), and increases in tropical diseases (Husnina et al., 2019).

The dearth of data for tropical forest regions other than the Amazon makes assessments of deforestation-related changes in temperature, precipitation and streamflow difficult ( ''high confidence'' ), and hampers estimates of tropical forest ecosystem health, biodiversity loss and vulnerability to current and future climatic and other pressures ( ''high confidence'' ). There is, hence, a strong need for increased investment in relevant data and research to narrow the knowledge gaps (Davison et al., 2021). Nonetheless, conclusions based on a newly developed tropical vulnerability index synthesising remotely sensed land use and climate information indicate that forests in the Americas are already reaching critical levels to multiple stressors, while forests in Asia reveal vulnerability primarily to land-use change and African forests still show relative resilience to climate change (Saatchi et al., 2021).

<div id="CCP7.4" class="h1-container"></div>

<span id="ccp7.4-social-economical-vulnerabilities-of-indigenous-peoples-and-local-communities-living-in-tropical-forests"></span>