Local conditions and policy design determine whether ecological compensation can achieve No Net Loss goals

Halting biodiversity loss and securing ecosystem services are fundamental challenges. Industrial development, while crucial for economic goals, puts immense pressure on ecosystems. Many nations employ ecological compensation policies to mitigate these negative impacts, aiming for "No Net Loss" (NNL) of biodiversity and ecosystem services. These policies follow a mitigation hierarchy: avoid, minimize, remediate, and offset losses. Hundreds of such policies exist globally, yet their effectiveness remains uncertain. Some even appear to facilitate ongoing biodiversity loss and damage ecosystem services. To make ecological compensation a cornerstone of global conservation, understanding its successes and failures is vital. Two key complicating factors are the vast variation in policy design and the substantial variability in biodiversity gains across different locations. Some policies focus on restoration, while others prioritize protecting existing biodiversity. The effectiveness of restoration depends on local conditions, such as the proximity to healthy ecosystems. Developing realistic counterfactual scenarios (what would happen without intervention) is challenging due to varying biodiversity trajectories. This study investigates the interaction between policy design and local conditions to determine their influence on policy performance—how close compensation comes to achieving NNL goals. Previous research has explored specific policies and outcomes, but this study is the first to systematically examine a common set of policy settings across multiple case studies with varying local conditions. We analyze 18 different policy options, representing combinations of area-based approaches, biodiversity trades between development and compensation sites, and methods for prioritizing compensation.

The introduction adequately summarizes the existing literature on ecological compensation policies, highlighting the inconsistencies in their effectiveness and the challenges in evaluating their performance. It points out the lack of systematic studies comparing policy designs across diverse contexts, positioning this research as a crucial contribution to the field. The paper cites relevant works on the mitigation hierarchy, the different approaches to generating biodiversity gains (Averted Loss and Improvement), the challenges of setting appropriate multipliers for compensation, and the importance of considering counterfactual scenarios and local conditions. The reference to hundreds of existing policies worldwide emphasizes the scale of the issue and the need for improved understanding.

This study uses spatial simulation models to quantify the impacts of different ecological compensation policies on biodiversity and ecosystem services. The researchers employed a four-step modeling approach: 1. **Quantify impacts of development on biodiversity:** They determined the extent and spatial distribution of future regulated development requiring compensation, varying across four case studies (Australia, Brazil, Indonesia, and Mozambique). Methods for determining development footprints varied—land-use change simulation models were used for some studies, while maps of proposed projects were used for others. 2. **Allocate compensation to the landscape:** Eighteen policy design options were tested, combining two approaches to biodiversity gains (Averted Loss and Improvement), four types of biodiversity trades (Out-of-Kind, In-Kind, Trading-up: Rarity, and Trading-up: Additional Gains), and three prioritization methods (Outside PAs, Near PAs, Within PAs). A Dinamica EGO model allocated compensation based on these constraints, aiming for the protection of existing vegetation (Averted Loss) or the restoration of cleared land (Improvement). 3. **Simulate counterfactual losses and gains:** Land-use change models, calibrated and validated using historical data, simulated unregulated biodiversity losses and gains in the absence of regulated development and compensation. This provided a baseline against which to measure the effectiveness of the policies. 4. **Quantify compensation impacts and outcomes:** The impacts of each policy were quantified by comparing counterfactual scenarios with scenarios including regulated development and compensation. The study measured impacts on native vegetation extent (as a biodiversity proxy) per vegetation type and on two ecosystem services: carbon storage and sediment retention. InVEST and other relevant models were used for this quantification. Four case studies, each with different local conditions, were selected to assess the influence of context on policy performance. The study's key assumptions include a 50% restoration success rate for Improvement approaches and the complete removal of biodiversity by development. The methodology also acknowledges potential limitations arising from data limitations and model assumptions. The paper's transparency in describing data sources, model details, and assumptions is commendable.

The key findings of this study demonstrate that none of the 18 investigated compensation policy designs achieved No Net Loss (NNL) of native vegetation extent (used as a biodiversity proxy) in any of the four case studies. Two crucial factors constrained NNL achievement: * **Land availability for compensation:** Insufficient land for either protecting existing vegetation (Averted Loss) or restoring cleared land (Improvement) limited compensation potential. This limitation was most pronounced in East Kalimantan, where the land required for restoration was double the available area. * **Counterfactual biodiversity losses and gains:** The extent of unregulated vegetation clearing (counterfactual losses) and natural recovery (counterfactual gains) significantly influenced policy performance. Averted Loss approaches performed better when counterfactual losses were substantial, while Improvement approaches were negatively affected by high rates of natural recovery. Policy design also played a crucial role: * **Approach to biodiversity gains:** Improvement approaches sometimes outperformed Averted Loss, but this depended heavily on the chosen multipliers (the amount of compensation required per unit of development impact) and the assumption of 50% restoration success. The study highlights the need for accurate estimates of counterfactual losses and gains in setting multipliers. * **Type of biodiversity trade:** Trading-up (prioritizing the protection or restoration of high-conservation-value areas) performed well, especially when using Averted Loss in areas with significant counterfactual losses. * **Prioritization of compensation:** Prioritizing compensation outside protected areas was generally more effective than near or within protected areas, although exceptions existed in Cabo Delgado due to the spatial distribution of protected areas relative to development pressures. Even the best-performing policies failed to significantly reduce regional biodiversity declines because the policies regulated only a subset of development sectors. Expanding policy scope to include more sectors would require even more land than is currently available for compensation. The study found that some policies achieved NNL for specific vegetation types but not overall. The performance of compensation policies for ecosystem services (carbon storage and sediment retention) varied widely and depended on their impact on biodiversity and the links between biodiversity and ecosystem services at both development and compensation sites. Some policies performed better for ecosystem services than for biodiversity, while others showed the opposite trend, emphasizing the need to address multiple goals explicitly.

The study's findings underscore the limitations of current ecological compensation policies in achieving NNL goals. While the study demonstrates the impact of policy design and local conditions, it also highlights the inherent limitations of compensation approaches when facing extensive biodiversity losses from unregulated development. Even perfectly implemented compensation strategies (as assumed in this study) have only a small influence on slowing regional biodiversity loss, as demonstrated by a less than 10% reduction in region-wide vegetation loss in three out of four case studies. The limited scope of existing policies only addresses a fraction of the biodiversity loss drivers. Expanding policy scope is theoretically possible, but practically constrained by land availability. The study suggests that impact avoidance—preventing biodiversity loss in the first place—is essential for halting biodiversity decline and achieving NNL goals, particularly once compensation opportunities are exhausted. The study contributes significantly to the understanding of ecological compensation, emphasizing the importance of considering local conditions and the limitations of current policies in achieving large-scale conservation goals. The findings have significant implications for policy design, urging a shift from a reliance on compensation towards stricter regulation to prevent biodiversity loss and promoting a target-based approach for ecological compensation.

This research demonstrates that current ecological compensation policies, even under ideal conditions, are insufficient to achieve No Net Loss (NNL) of biodiversity at regional scales. The study's findings highlight the critical roles of land availability and counterfactual biodiversity changes in limiting the effectiveness of compensation. Policy design, including the selection of appropriate multipliers and approaches to biodiversity gains and trades, significantly influences outcomes. The limited scope of most policies further constrains their ability to mitigate broader biodiversity losses. The study strongly advocates for a greater emphasis on impact avoidance as a crucial step in achieving NNL goals. Future research should focus on developing more sophisticated models that incorporate factors like habitat condition, temporal dynamics, and leakage effects to provide more accurate assessments of compensation policies and guide the design of more effective strategies.

Several limitations affect the generalizability of the study's findings. First, the use of native vegetation extent as a biodiversity proxy oversimplifies the complexity of biodiversity. Second, the assumption of complete biodiversity removal by development may overestimate impacts, particularly for development types that do not require complete land clearing. Third, the assumption of a 50% restoration success rate introduces uncertainty, although sensitivity analyses explored alternative success rates. Fourth, data limitations and model uncertainties affect the accuracy of the results, although this uncertainty was partially quantified. Finally, the study didn't assess potential biodiversity losses due to leakage or displacement of unregulated development from compensation sites. The study's scope, focusing on broad-scale analysis, limits the ability to provide specific policy recommendations for individual case studies.