1.3.2.3 Temperate forests

Temperate forests cover around 767 million hectares (16 per cent of the global forest area) and represent 34 per cent of global carbon sinks, storing around 119 GtC (Hansen et al., 2010; Pan et al., 2011) (Figure 1.3.6). In this report, we only consider temperate forests as defined in Figure 1.3.1. Mediterranean forests are covered under Drylands [see 1.3.2.5].

In most regions their spatial cover is highly fragmented following a long history of human land-use and forestry practices. In fact there are only a few temperate forests which are considered ‘intact’ primary forest (Potapov et al., 2017; Sabatini et al., 2021) and the vast majority are managed by humans using vastly varying forest management techniques and intensities.

Current managed temperate forests are often monocultures or mixtures of few tree species with relatively low biodiversity and structural diversity, optimised for high timber yields and certain wood features established under the assumption of stable climate and environmental conditions (instead of optimised for long-term forest resilience).

Evidence for tipping dynamics

In recent years temperate forests globally have suffered enormous damages and losses caused by extreme heat waves and droughts in combination with secondary effects like insect outbreaks and fires (Allen et al., 2010; Buras et al., 2019; Senf et al., 2020; Zhang et al., 2021; Carnicer et al., 2021; Benyon et al., 2023; Forzieri et al., 2022). As many temperate forests are effectively plantations for wood production in most parts of the world, those impacts often occurred in a similar synchronised manner on regional scales. Embedded in landscapes dominated by human land use (segregated by roads, crops, power lines, etc.), many temperate forests feature reduced connectivity and hence less exchange of species or genetic material, which reduces resilience (Sabatini et al., 2021). 

More importantly, the extremely low diversity reduces the forest’s ability to cope with stress through mechanisms such as portfolio insurance effects or complementarity (Billing et al., 2020). Portfolio insurance effects refer to the idea that having a diverse portfolio of species can help protect the forest against stressors by spreading the risk among different species. Complementarity refers to the idea that different species work together in a complementary way to improve the overall functioning of the ecosystem. However, when there is low diversity, these mechanisms may not be as effective and hence a potential tipping of temperate forests might also be more abrupt than in natural systems. Still, it must be noted that effective support from forest management (by regenerating an area through planting or supporting natural regeneration) can in principle also alleviate some of the pressures that natural systems face. 

Besides the clear devastating signals of temperate forest damage and dieback, past assessments have had difficulties classifying temperate forests as tipping systems. In a review by (Thom, 2023) many temperate forest ecosystems were identified as resilient and/or resistant to increasing disturbance regimes and unlikely to shift towards alternative states in the very near future at large scale. However, drastic changes under intensifying future pressures such as climate change cannot be ruled out. In accordance with these findings, the recent assessment of (Armstrong McKay et al., 2022) has categorised temperate forests as an uncertain potential regional impact tipping system.

So far self-amplifying feedbacks in temperate forest dieback were described for more localised landscape-scale stressors like bark beetle attacks and fire in the Boreal forest section (see 1.3.2.2 and Box 1.3.2) (Hlásny et al., 2021; Fettig et al., 2022). On larger spatial scales it remains less clear whether temperate forests might feature self-amplifying feedbacks strong enough to induce tipping behaviour. However, just as in the tropical zone, the principles of cascading moisture recycling also apply to temperate forests. Any loss of forest cover reduces atmospheric moisture supply, hence reducing precipitation downwind and increasing sensible heat, which can amplify drying and warming in the affected areas (Pranindita et al., 2021). The average net cooling effect of temperate forests compared to grassland was found to be 1-2°C, with maxima of up to 5°C (Zhang et al., 2020). A recent study integrating data and modelling results reports continental-scale cooling effects of regrowing temperate forests on abandoned agricultural areas (Huang et al., 2020). 

Related to this, cloud formation probability was found to be higher above forests in comparison to other land cover types in the temperate region (Teuling et al., 2017). Therefore, recent forest damages could have decreased cloud cover during recent droughts and heatwaves further intensifying these events. Furthermore, soil moisture-atmosphere feedbacks related to droughts and heatwaves were reported for the temperate zone (Seneviratne et al., 2010; Jaeger and Seneviratne, 2011) and could indicate that droughts might self-propagate in space and time (Schumacher et al., 2022). A recent study for the US west coast suggests cascading effects of soil moisture and biomass during a multi-year drought (Au et al., 2023). The recent large-scale forest damages and losses in the temperate zone (Senf et al., 2020; Lloret and Batllori 2021) could mark the beginning of self-amplifying and potentially self-sustaining feedbacks, but further work is required to confirm this.

The most important mediator between soil moisture and the atmosphere is vegetation, and forests especially stand out since they access water in great depths via their root systems (Sakschewski et al., 2021; Singh et al., 2020; Fan et al., 2017). Hence, larger-scale forest damage or loss means losing this mediator, further decreasing atmospheric moisture supply and downwind rainfall. This becomes particularly significant when, during droughts, precipitation becomes increasingly dependent on water evaporated from land or transpired by vegetation due to altered atmospheric patterns (Pranindita et al., 2021). 

Additionally, local mechanisms or secondary effects could increase the likelihood of nonlinear responses, thereby increasing the probability of reaching tipping points. For instance in a more open forest or simply due to warmer and drier conditions at the forest floor, fire occurrences and intensities can easily increase. Moreover, the suppression of forest regeneration can occur due to the invasion of highly competitive light-demanding plant species, forming ecosystems which potentially transpire less moisture back to the atmosphere. 

In combination with reduced resilience and resistance due to human interferences, abrupt large-scale damage and dieback of temperate forests is conceivable. Early warning signals in satellite-derived biomass data hint towards such a destabilisation (Forzieri et al., 2022). Yet, large-scale tipping behaviour in temperate forests is not proven. If at all, this will certainly be region-specific and recent forest damage will illuminate such potential feedbacks in the near future.

Assessment and knowledge gaps

It is uncertain if temperate forests have strong enough self-amplifying feedbacks like the Amazon rainforest and boreal forest to result in tipping, hence there is no evidence for larger-scale tipping and confidence is low. There is, however, a lot of evidence and medium confidence for abrupt changes with changing disturbances regimes.

Human forest management practices may have made temperate forests less resilient and therefore more susceptible to abrupt changes, but improved management can assist resilience and adaptation to climate change. Based on the impacts of current extreme events on temperate forests, it can be inferred that an increase in the intensity and/or frequency of such events could severely threaten existing forests in many areas, even without further climate change (Senf et al., 2020; Lloret and Batllori, 2021). The potential feedback to the water cycle requires further investigation. In particular, modelling studies should fully account for extreme events such as droughts, heatwaves and other important disturbances, their increasing frequencies and intensities as well as their potential impact on simulated vegetation and the resulting land-atmosphere feedbacks (Kolus et al., 2019).