Current trends suggest most Asian countries are unlikely to meet future biodiversity targets on protected areas

Area-based sub-target

We found that 13.2% of Asian terrestrial landscapes were covered by PAs by the target date for Aichi 11 based on our in-country sources. However, it was 17.4% lower based on WDPA data (10.9%). The average increase in coverage across Asia during the 2010s was 0.4% ± SE 0.1% per year. PA coverage at the level of individual countries increased from a mean 11.1% in 2010 (SE = 1.4%) to 14.1% by 2020 (SE = 1.8%) based on our in-country sources, which was 16.5% higher than WDPA data (12.1 ± SE 1.6%). However, these overall figures concealed considerable country-level and sub-regional heterogeneity.

A total of 8,673,433 km² across 10 countries, equaling 19.6% of Asian terrestrial landscapes was managed as hunting concessions, governed by governments, communities or private sectors, but these areas have not been included in the countries’ report to the Protected Planet Initiative databases. Most of these areas are locally important in terms of biodiversity conservation and local socioeconomic outcomes which may qualify them as examples of “other effective area-based conservation measures” (OECMs). The increase in area-based conservation coverage represented by these areas, above the current Protected Planet Initiative statistic, ranged from 0.2% (Iran) to 41.4% (Russia). With that update incorporated, a total of 32.9% of Asian terrestrial landscapes are under protection, either as protected areas or hunting concessions (potentially as one type of OECMs).

We found that 40% of Asian countries met a target of 17% coverage for PAs by 2020 based on our in-country sources, mainly in East and some South Asia, whereas West and Central Asian countries had generally not achieved this target (Figs. 1 and 2). We did not find any statistically significant association between the proportions of highly at-risk (CR/EN) mammalian species range outside PAs and the % PA extent in 2020 (β = −0.22 ± SE 0.15, t = −1.51, P = 0.14 in a Generalized Linear Model). The highest proportions of the highly at-risk (CR/EN) mammalian species range outside PAs were seen in West (β_{CR/EN_outsidePA} = 1.77 ± SE 0.46, t = 3.86, P < 0.001) and East Asia (β_{CR/EN_outsidePA} = 2.07 ± SE 0.61, t = 3.39, P < 0.001), respectively. Indeed, the lower-diversity sub-region of East Asia showed the greatest level of protection, followed by higher-diversity Southeast Asia.

When we considered the pattern of terrestrial PA expansion for 2010-2020 (rather than the final coverage achieved), two countries in Southeast Asia (Myanmar and Thailand) showed small decreases in PA coverage (PA downgrading, degazettement and downsizing) whereas Indonesia and Cambodia showed significant (>10%) expansion. In South and West Asia, most countries showed an increase in area, with Bhutan and Qatar gaining >10%, but Kuwait lost area. In East Asia, all countries showed at least some PA expansion (South Korea and Japan by >10%) whereas in Central Asia, almost no change was seen. It is also noteworthy that between 2010 and 2015, agricultural lands increased by 2.0% across the continent, averaging 0.51 ± SE 0.03% per year at country level, although 18 counties (45.0%) had agricultural land loss, mainly in West and Central Asia (12 out of 18 countries with agricultural land loss; Fig. 2).

In our attempt to model the variation in achievement of area-based target (% PA extent), we found a single model with a ΔAICc weight of 1.0 (R²_adj = 0.66; Table 1). There was no evidence to reject the null hypothesis that the model fits well (P = 0.99). This model included the predictors % agricultural extent in 2015, % PA extent in 2010, and sub-region (Table 1). Specifically, the coefficients suggested that countries with greater PA extent in 2010 and a smaller percentage of agricultural lands in 2015 were more likely to achieve higher percentage of PA extent by 2020 (β_PAExtent2020 = 0.58 ± SE 0.10, t = 5.74, P < 0.001 and β_{Agriculture2015} = −0.36 ± 0.11, t = −3.25, P = 0.003; Fig. 3). In the sub-regional effect, the smallest and largest % PA extents in 2020 were found in West (P = 0.01) and Southeast Asia (P = 0.08) respectively (Fig. 2).

For PA growth rates, we found that with the current annual growth rate (r) of 0.03 for 2010–2020, Asia can achieve only 18.6% of PA extent by 2030. A 2.5 times larger r (0.08) would be needed to meet the 30% target by 2030 for the entire continent. The continent as a whole is therefore projected to miss a putative 30% areal target by a considerable margin. On a sub-regional level, all regions except Central Asia are projected to miss the putative 30% target, with West and South Asia as the most underperforming regions, projected to achieve 11.3 and 10.0% coverage respectively by 2030. Achieving the post-2020 Global Biodiversity Framework’s target of at least 30% of the planet protected would need 2.4 times faster PA growth for East and Southeast Asia, 4.4 times faster in West Asia and 5.9 times faster in South Asia.

Ecological representativity sub-target

For ecological representativity, we found that 73.1% (n = 196) of 268 terrestrial ecoregions in Asia had less than 17% coverage inside the PA network (Fig. 4). Nonetheless, apart from Socotra Island xeric shrublands, no terrestrial ecoregion in West Asia achieved ≥17% of PA coverage. Similarly, the overlap between the PA network and the ecoregion did not exceed ≥17% in central Asia, except three ecoregions of Altai Alpine meadow and tundra, Ural montane forests and tundra and Pamir alpine desert and tundra.

In our sub-analysis of how well PA systems covered threatened species, we found that for 241 highly at-risk (CR/EN) mammalian species across Asia, a mean of 84.4% of their ranges fell outside the PA network (SE = 2.3%). A single model was selected for ecological representativeness with a weight of 0.71, without any evidence to reject the null hypothesis that the model fits well (P = 0.96). The model included only Region as factor (R²_adj = 0. 11; Table 2) but did not show any evidence for geographic differences across Asian regions (P > 0.05).

For the coverage of highly at-risk (CR/EN) mammalian species, a single statistical model was also selected, with non-significant deviance goodness of fit (P = 0.83), which included only the % PA extent by 2020 and Region as predictors (R²_adj = 0. 27). Although there was no evidence for association between the % PA extent by 2020 and the coverage of threatened species (β_PAExtent2020 = −0.23 ± SE 0.15, t = −1.57, P = 0.13). However, the coverage of threatened species varied geographically, with high intercept differences for East Asia (β_EastAsia = −0.23 ± SE 0.15, t = −1.57, P = 0.13), implying the largest median of range of highly at-risk (CR/EN) mammalian species outside the current network of PAs within each country.

PA management effectiveness sub-target

For the level of PAME assessment, we found that out of 22781 PAs within the 40 studied Asian countries, only 7.0% have been assessed based on PAME criteria (n = 1599), averaging 17.4% ± of PAs per country (SE = 2.5%). Israel, Japan, Lao, Bahrain, Oman and Qatar had no PA assessed based on the PAME criteria while over 1/3 of PAs in Indonesia, Cambodia, Bhutan, Jordan, Nepal, Turkey, Singapore and the UAE were PAME assessed. When modeling the level of PAME assessment, three best supported models were averaged (Table 3), with the averaged model including GDP2019, % PA extent 2020 and the Region as predictors. The averaged model coefficients would be non-significant under a hypothesis-testing approach (β_GDP2019 = −0.18 ± SE 0.12, t = 1.47, P = 0.14 and β_PAExtent2020 = −0.15 ± SE 0.11, t = 1.31, P = 0.19). Similarly, there was no evidence for the association between the ratio of PAs with PAME and Asian regions (P > 0.05).

Source: Ecology - nature.com