The Deer and the Tiger, the Forest and the Carbon

Healthy, verdant forest ecosystems are indicated by an intact complement of highly abundant vegetation along with diverse and abundant populations of herbivore species and their carnivore predators (Wang et al. 2025). Keeping forest ecosystems verdant is considered vital to ensuring planetary resilience to climate change (Watson et al. 2018). This is because the highly abundant vegetation takes up atmospheric carbon that is then stored in vegetation biomass and in soils of those ecosystems (Pan et al. 2024).

But what would happen to climate change resilience if forest ecosystem intactness was disrupted by, say, the loss of predators? The answer depends on how intactness is sustained. Ecological science has two general views on this (Schmitz et al. 2018). One—the bottom-up control view—holds that ecosystems are verdant because vegetation abundance, which is strongly determined by soil nutrient and moisture levels, supports but limits the abundance of herbivore populations, and in turn, populations of their carnivore predators. In this view, climate change resilience would not be disrupted by predator losses (or herbivores for that matter) because animal abundance and diversity do not drive vegetation abundance. The other—the top-down view—holds that predators, by virtue of limiting the abundance of their herbivore prey, keep ecosystems verdant by preventing herbivore overexploitation of vegetation. In this view, climate change resilience would be disrupted by the loss of predators because they ultimately drive vegetation abundance. An important and challenging research frontier for both ecological and global change science is resolving which view of control best explains the climate resilience of verdant forest ecosystems, especially across vast landscapes over which large animals live and roam (Schmitz et al. 2018). This challenge is compounded by variation in biophysical conditions across those vast landscapes because the amount of carbon captured and stored among geographic locations becomes highly dependent upon biophysical context, including climatic conditions, the kinds, diversity, and abundances of plant and animal species, and the nutrient contents and physical properties of soils (e.g., Sobral et al. 2017; Schmitz et al. 2018; Schuldt et al. 2023).

In newly published research in Global Change Biology, Roberts et al. (2025) address this formidable challenge to reveal how varying abundance and outright loss of a large predator—the tiger (Panthera tigris)—and the abundance of its “deer” (i.e., ungulate) prey species are related to the capture and storage of carbon in forests across the tiger's geographic range throughout Asia. This vast landscape has myriad biophysical dimensions that create much context dependency. It contains four broadly different forest ecosystem types, including boreal, temperate, subtropical dry, and subtropical moist. The ecosystems are arrayed across different elevations from lowland to montane and across different wet to dry regimes. They harbor many varieties and abundances of plant and herbivore prey species.

Resolving the role of tigers in forest climate resilience requires comparing carbon dynamics in places where tigers are present with similar environmental conditions where they are absent (Schmitz et al. 2018). The hallmark scientific approach would be to conduct a manipulative Case (tiger-present target forest)–Control (tiger-empty forest) experiment that pairs biophysical conditions of forest locations and systematically excludes tigers from half of them, replicated across the different forest ecosystem contexts across the vast landscape (Schmitz et al. 2018). This is altogether logistically impossible. Roberts et al. (2025) overcame this critical limitation with the next best approach (Schmitz et al. 2018): they amassed a large multivariate data set gathered from satellite remote sensing and extensive on-the-ground sampling across landscape locations where tigers have been persistently present versus similar landscape conditions where they are absent—their loss arising from known historical extirpation by humans. The amassed data included measures of carbon in vegetation biomass and soil, net ecosystem carbon exchange, and aspects of context dependency including forest ecosystem type, regional climate, and human disturbance. These are exactly the kinds of data used in the conventional accounting of carbon capture and storage in forest ecosystems (Pan et al. 2024). As well, they accounted for context dependency in top-down control due to variation in deer diversity and abundance, which according to classical theory (Polis and Strong 1996) should be stronger in more linear food chains (i.e., deer diversity is low) and dissipate along the branching network of linkages in food webs with higher prey diversity. There was sufficient replication within and among forest ecosystem types to control for underlying spatial context dependency in these biophysical factors in both tiger present and absent sites. Hence, Roberts et al. (2025) were able to undertake extensive direct and counterfactual analyses to fully and robustly interrogate the veracity of the top-down view that tigers play a role in controlling forest carbon dynamics across their geographic range.

The short answer to the question about the consequences of predator loss is that it mattered. Tiger presence was generally linked to higher vegetation carbon stocks and higher net carbon exchange (i.e., forest carbon uptake exceeded carbon emissions) among forest types. But the longer answer is that there was also nuance due to context-dependency. Certainly, bottom-up control prevailed across all forests due to geographic variation in biophysical properties that created variation in vegetation biomass and carbon exchange and storage among the forest ecosystems. But top-down effects of tiger presence also prevailed, resulting in higher vegetation and soil carbon stocks among most contexts than in the absence of tiger, the exceptions being tropical dry forest and forests in lowlands and especially wetlands. These exceptions extended to net ecosystem carbon exchange, with one context (tropical swamp forest) even switching from being a slight net carbon sink to a carbon source in the presence of tigers. Tiger presence was impactful in forests with low to intermediate vegetation biomass but not high biomass, suggesting that the biomass dense forests might be completely bottom-up controlled. Yet prey diversity also seems to have played a hand by mediating tiger effects. Forests with low to intermediate vegetation biomass had less deer biomass and diversity than did high vegetation biomass forests. Hence, a second reason for weak, if any, tiger control in biomass dense forests is that the top-down effects may have dissipated along the many food web linkages, suggesting again that both top-down and bottom-up control persisted across contexts but with different relative importance.

In 1967 wildlife ecologist George Schaller published a seminal treatise The Deer and Tiger in which he provided foundational scientific understanding about the ecological interplay between tigers and their key prey species. It further described how those relationships were being eroded due to rural land use change for agriculture, persecution of tigers out of fear for human safety and loss of wellbeing, and exploitation for body parts that had perceived medicinal value—problems which continue to this day (Roberts et al. 2025). Schaller's book was published at a time when ecological science had barely started to imagine that predator–prey dynamics could cascade to impact the properties and functions of entire ecological communities, let alone the kinds of ecosystem processes such as the carbon cycle that also have implications for human wellbeing. In retrospect it should have been a portent of ramifying ecosystem impacts. But at the time the scientific community was ill-equipped conceptually to even recognize this as a bigger issue because it had yet to embrace a holistic perspective on what it means to sustain verdant forest ecosystems. The portent is still unheeded in many areas of climate change science today, even while scientific understanding of the role of animals in controlling the carbon cycle has advanced considerably (Schmitz et al. 2018). By expanding the scientific story of the Deer and the Tiger to the story of the Forest and the Carbon, Roberts et al. (2025) have provided a leap in scientific understanding about how broadly impactful an animal species can be.

These findings are sure to excite those working on the frontlines of tackling biodiversity loss and climate change by showing how both looming problems can begin to be solved together. There is, however, a risk that arguments about the carbon benefits of tigers will be used to justify headlong action to restore this iconic and endangered species everywhere across its geographic range. Here the deeper, and critical, lesson of the Roberts et al. (2025) article is that by deliberately addressing context-dependency it provides the kind of sober analysis that has been recently called for to temper policy and conservation from overpromoting animal restoration and conservation as a universal win-win for mitigating biodiversity loss and climate change together (Burak et al. 2024). The science is clear that tigers can have impactful effects on forest ecosystem carbon storage. But most importantly, Roberts et al. (2025) have provided unprecedented scientific insight into which forest ecosystems across the tiger's vast geographic range should be considered candidates for tiger population conservation and restoration for carbon capture and storage; and which forests definitely should not. The research is exemplary for showing how to advance an evidence-based approach that can be applied in the service of restoring and sustaining verdant forest ecosystems for the purpose of maintaining resilience to climate change.

The author declares no conflicts of interest.

This article is a Invited Commentary on Guangshun Jiang et al., https://doi.org/10.1111/gcb.70191.