Climate co-benefits of tiger conservation

Abstract

Biodiversity conservation is increasingly being recognized as an important co-benefit in climate change mitigation programmes that use nature-based climate solutions. However, the climate co-benefits of biodiversity conservation interventions, such as habitat protection and restoration, remain understudied. Here we estimate the forest carbon storage co-benefits of a national policy intervention for tiger (Panthera tigris) conservation in India. We used a synthetic control approach to model avoided forest loss and associated carbon emissions reductions in protected areas that underwent enhanced protection for tiger conservation. Over a third of the analysed reserves showed significant but mixed effects, where 24% of all reserves successfully reduced the rate of deforestation and the remaining 9% reported higher-than-expected forest loss. The policy had a net positive benefit with over 5,802 hectares of averted forest loss, corresponding to avoided emissions of 1.08 ± 0.51 MtCO₂ equivalent between 2007 and 2020. This translated to US$92.55 ± 43.56 million in ecosystem services from the avoided social cost of emissions and potential revenue of US$6.24 ± 2.94 million in carbon offsets. Our findings offer an approach to quantitatively track the carbon sequestration co-benefits of a species conservation strategy and thus help align the objectives of climate action and biodiversity conservation.

Fig. 1. The locations and avoided carbon emissions of analysed tiger reserves.

Top left: map of India with the reserves analysed in the study (boundaries from OSM). Light grey donor reserves are protected areas with tiger presence that did not undergo enhanced conservation policy and were used to generate synthetic counterfactuals for tiger reserves. Dotted bounded boxes represent geographical areas used for donor matching. To ensure robust significance testing, more than 20 donor reserves or placebos were needed per geographical grouping. Therefore, contiguous tiger conservation landscapes were combined into the following groups: Shivalik–Gangetic, Central India, Eastern Ghats and Sunderbans regions (A); Western Ghats (B); and Northeast Hills and Brahmaputra region (C). For significance testing, we used a two-sided Fisher’s exact test to compare the ratios of pre-intervention and post-intervention mean squared prediction errors between the treated synthetic reserve and placebo units for each tiger reserve (see Extended Data Fig. 1 for unadjusted P values of tiger reserve with significant effects). Green reserves represent tiger reserves that exhibited significantly avoided deforestation while yellow reserves are treated reserves where the observed deforestation was significantly higher than the synthetic counterfactual (P < 0.05). Dark grey reserves represent tiger reserves that yielded insignificant results (see Supplementary Table 2 for list of P values). Top right: total avoided emissions per reserve. Data are presented as mean avoided emissions ± uncertainty values (based on mean cumulative standard errors reported by ref. ) in ktCO₂e. Error bars were derived by multiplying avoided deforestation, an emissions factor of 3.67 and mean values of the cumulative standard errors in predictions of above- and belowground carbon biomass densities reported by ref. for each reserve. Reserves are numbered and colour coded to indicate locations on maps. Bottom, zoomed-in geographical zones with the number of tiger reserves and donor reserves used for deriving counterfactuals and statistical significance included for each zone. Map boundaries from OpenStreetMap under a Creative Commons license CC BY-SA 2.0.

Fig. 2. Trend lines for cumulative forest loss in tiger reserves that exhibited significant avoided deforestation.

All reserves displayed here exhibited significant avoided deforestation (unadjusted P < 0.05) based on a two-sided Fisher’s exact test to compare the ratios of pre-intervention and post-intervention mean squared prediction errors between placebo and treated units (see Methods for details). Significance levels (unadjusted P values) were reported for each synthetic counterfactual in the displayed plots. Only reserves that exhibited avoided forest loss of more than 10 ha are displayed here (see Extended Data Fig. 1 for trend lines for all 11 tiger reserves with significant avoided forest loss in the study period). The dotted pink line represents the cumulative forest loss for the synthetic control model, whereas the dotted grey line represents observed deforestation in hectares. The vertical dashed line represents the year of implementation of the enhanced conservation policy. For each of these reserves, the synthetic control line closely tracks the observed cumulative forest loss values before the intervention.

Fig. 3. Trend lines for cumulative forest loss in tiger reserves that exhibited significantly higher than anticipated deforestation.

All reserves displayed here exhibited significant results (unadjusted P < 0.05) based on a two-sided Fisher’s exact test to compare the ratios of pre-intervention and post-intervention mean squared prediction errors between placebo and treated units (see Methods for details). Note the observed deforestation values were higher than the synthetic counterfactual in the case of these reserves, suggesting higher than anticipated forest loss. Significance levels (unadjusted P values) were reported for each synthetic counterfactual in the displayed plots. The dotted pink line represents the cumulative forest loss for the synthetic control model, whereas the dotted grey line represents observed deforestation in hectares. The vertical dashed line represents the year of implementation of the enhanced conservation policy. For each of these reserves, the synthetic control line closely tracks the observed cumulative forest loss values before the intervention.

Extended Data Fig. 1. Trend lines for cumulative forest loss in Tiger Reserves where the conservation policy exhibited significant effects on deforestation.

Cumulative forest loss for Tiger Reserves that exhibited significant results (unadjusted p-value < 0.05) based on a two-sided Fisher’s exact test to compare the ratios of pre-intervention and post-intervention mean squared prediction errors between treated reserves and placebo units (see Methods section for more details). Significance levels from placebo testing have been reported for each synthetic counterfactual in the displayed plots. Overall, 15 out of the 45 reserves exhibited significant effects after Tiger Reserves with anticipation effects ruled out (See Supplementary Figs. 5, 6 and Supplementary Table 3). Of these 15 Tiger Reserves, 11 demonstrated avoided deforestation. The remaining four reserves demonstrated higher than anticipated forest loss (highlighted in the dashed red box). The dotted pink line represents the cumulative forest loss for the synthetic control model while the dotted grey line represents observed deforestation in hectares. The vertical dashed line represents the year of implementation of the enhanced conservation policy. For each of these reserves, the synthetic control line closely tracks the observed cumulative forest loss values before the intervention.