Search

Ecological corridors are crucial for preserving biodiversity in human-impacted areas, but previous work has focused on landscape connectivity rather than ensuring that connected areas are suitable for migrating species. Our study developed a framework integrating landscape similarity assessment to prioritize corridor protection and construction. Taking Nanjing Metropolitan Area in 2020 as a case, we identified ecological sources from the perspectives of function and structure. The minimum cumulative resistance (MCR) model and gravity index were then employed to detect ecological corridors and assess connectivity. The similarity index was subsequently applied to assess the landscape similarity between ecological sources, which was quantified from three aspects of composition, as well as configuration and physical geography. Final classification and prioritization of corridors for protection and construction were based on the precedence matrix method. Our results showed that 58 ecological sources were potentially connected by 103 corridors. The 39 corridors combining high connectivity and high similarity were assigned the highest priority for protection, and the 27 corridors with low connectivity but high similarity were assigned to the category with high construction priority. Our findings advance theoretical and practical understanding of ecological corridors by demonstrating that high priority corridors can be identified for focused conservation efforts.

The contradiction among human being, resources, and environment has become a significant obstacle to achieving sustainable development, especially in rural areas subject to the spillover of urban development elements. With the immense strain of resources and environment, it is critical to assess whether human activities fall within the carrying capacity range of a natural ecosystem in a rural system. Taking the rural areas of Liyang county as an example, this study aims to assess the rural resource and environmental carrying capacity (RRECC) and diagnose its critical obstacles. Firstly, a social–ecological framework focusing on human–environment interaction was employed to construct the RRECC indicator system. Subsequently, the entropy-TOPSIS method was introduced to assess the performance of the RRECC. Finally, the obstacle diagnosis method was applied to identify the critical obstacles of RRECC. Our results show that the distribution of RRECC presents a spatial heterogeneity, with high- and medium–high-level villages primarily concentrated in the south of the study area, where there are abundant hills and ecological lakes. Medium-level villages are scattered throughout each town, and low and medium–low level villages are concentrated across all the towns. Moreover, the resource subsystem of RRECC (RRECC_RS) exhibits a similar spatial distribution to RRECC, while the outcome subsystem of RRECC (RRECC_OS) has a comparable quantity proportion of different levels to RRECC. Furthermore, the diagnosis results of critical obstacle vary between the town scale divided by administrative units and the regional scale divided by RRECC values. In detail, arable land occupied by construction is the main critical obstacle at the town scale, while the poor people in villages, people left-behind, and arable land occupied by construction are the main critical obstacles at the regional scale. Targeted differentiated improvement strategies for RRECC at regional scale from various p

Although compact urban form plays an important role in constraining emissions of carbon dioxide (CO2), the boundary for the impact of compact urban form on these emissions has nevertheless received little attention. We consequently applied the entropy weight method and several key landscape metrics to a dataset from 295 cities in China to quantify urban form compactness (UFC) between 2000 and 2015. The STIRPAT model then was employed to estimate the impact of UFC on CO2 emissions, and a panel threshold regression model was used to estimate threshold effects capable of limiting the impact of compact urban form on emissions. Although CO2 emissions increased sharply over the 15-year study period, a significant negative relationship between UFC and CO2 emissions was detected. Two thresholds of UFC were detected, and this allowed three categories to be differentiated: before the first threshold, between the two thresholds, and after the second threshold. These categories were respectively associated with no impact, strong impact, and weak impact of UFC on reduction of carbon emissions in the 295 cities. Carbon emissions reduction consequently becomes effective when the UFC exceeds the first threshold and effectiveness persists but at a reduced level when the UFC exceeds the second threshold. Further temporal analysis confirmed that an increasing number of mostly small- and medium-sized cities could constrain their future carbon emissions by adopting a compact urban form. Thus, government policies should emphasize UFC as a strategy to reduce CO2 emissions. Moreover, by defining the range of compact urban form that has the greatest impact on CO2 emissions, our study deepens the overall understanding of the influence of UFC on carbon emission reductions, so as to make contributions to the design of low-carbon cities.

Finding an optimal matching in a weighted graph is a standard combinatorial problem. We consider its semi-bandit version where either a pair or a full matching is sampled sequentially. We prove that it is possible to leverage a rank-1 assumption on the adjacency matrix to reduce the sample complexity and the regret of off-the-shelf algorithms up to reaching a linear dependency in the number of vertices (up to poly log terms).

We study a new non-stochastic federated multiarmed bandit problem with multiple agents collaborating via a communication network. The losses of the arms are assigned by an oblivious adversary that specifies the loss of each arm not only for each time step but also for each agent, which we call “doubly adversarial”. In this setting, different agents may choose the same arm in the same time step but observe different feedback. The goal of each agent is to find a globally best arm in hindsight that has the lowest cumulative loss averaged over all agents, which necessities the communication among agents. We provide regret lower bounds for any federated bandit algorithm under different settings, when agents have access to full-information feedback, or the bandit feedback. For the bandit feedback setting, we propose a near-optimal federated bandit algorithm called FEDEXP3. Our algorithm gives a positive answer to an open question proposed in (Cesa-Bianchi et al., 2016): FEDEXP3 can guarantee a sub-linear regret without exchanging sequences of selected arm identities or loss sequences among agents. We also provide numerical evaluations of our algorithm to validate our theoretical results and demonstrate its effectiveness on synthetic and real-world datasets.