The United Nations has declared 2021–2030 the decade on ecosystem restoration with the aim of preventing, stopping and reversing the degradation of the ecosystems of the world, often caused by the fragmentation of natural landscapes. Human activities separate and surround habitats, making them too small to sustain viable animal populations or too far apart to enable foraging and gene flow. Despite the need for strategies to solve fragmentation, it remains unclear how to efficiently reconnect nature. In this paper, we illustrate the potential of deep reinforcement learning (DRL) to tackle the spatial optimisation aspect of connectivity conservation planning. The propensity of spatial optimisation problems to explode in complexity depending on the number of input variables and their states is and will continue to be one of its most serious obstacles. DRL is an emerging class of methods focused on training deep neural networks to solve decision‐making tasks and has been used to learn good heuristics for complex optimisation problems. While the potential of DRL to optimise conservation decisions seems huge, only few examples of its application exist. We applied DRL to two real‐world raster datasets in a connectivity planning setting, targeting graph‐based connectivity indices for optimisation. We show that DRL converges to the known optimums in a small example where the objective is the overall improvement of the Integral Index of Connectivity and the only constraint is the budget. We also show that DRL approximates high‐quality solutions on a large example with additional cost and spatial configuration constraints where the more complex Probability of Connectivity Index is targeted. To the best of our knowledge, there is no software that can target this index for optimisation on raster data of this size. DRL can be used to approximate good solutions in complex spatial optimisation problems even when the conservation feature is non‐linear like graph‐based indices. Furthermore, our methodology decouples the optimisation process and the index calculation, so it can potentially target any other conservation feature implemented in current or future software.

Biological invasions need to be assessed as spatial processes, incorporating information on introduction sites, environmental suitability, dispersal parameters and negative impacts. The above allows us to develop risk maps, which are an important tool to determine the probability and consequences of an invasion in each area. In this study, we developed species niche-based distribution models for Vespa mandarinia and V. velutina, exotic species recently discovered in North America, being V. velutina a recognized invasive alien species that has caused enormous economic impacts in Europe. Species niche-based distribution models were used as a base map to determine the risk of establishment in conjunction with information related to the introduction, economic and biodiversity risk. The models developed in this study show environmental suitability for the establishment of these species in tropical and subtropical locations of North America. In Mexico, more than 50% of the ports are at high risk especially those located in the Gulf of Mexico. The biodiversity impact risk map for V. mandarinia shows that 57 protected areas are in regions with some risk of invasion and the V. velutina analysis shows 49 protected areas at potential risk. By implementing comprehensive surveillance and monitoring programs, integrating early detection and rapid response strategy and leveraging geographic information systems, Mexico can take proactive measures to mitigate the potential impacts of invasive species. These efforts will be crucial in protecting biodiversity, preserving ecosystems and safeguarding the country’s economy from the negative consequences associated with invasive species.

The implementation of agri-environmental practices (AEPs) is a key strategy to reach biodiversity and environmental objectives in agricultural landscapes, but their widespread application is often hampered by perceived trade-offs with crop production. However, the extent of these trade-offs remains poorly understood and has rarely been quantified in real-world case studies. Hence, our aim was to analyze trade-offs between crop yield, water quality and farmland biodiversity using an optimization approach for the spatial allocation of AEPs for a catchment in Eastern Germany. Potential AEPs were selected based on a co design approach with local stakeholders (stakeholder-based scenario) and were complemented by additional AEPs to reach current EU policies targets (policy-based scenarios). Consequences for crop production and environmental objectives were evaluated through spatially-explicit crop, water and biodiversity models. Contrary to common perception, we found that crop losses required to increase environmental objectives were marginal (maximum loss of 1.1% in the stakeholder based scenario). The implementation of AEPs even led to win-win outcomes for crop production and environmental objectives in over 20% of the Pareto-optimal solutions as compared to the status quo. These win-win outcomes resulted merely from biophysical effects as positive biodiversity feedbacks to agriculture were not included in our model. Spatial optimization of AEPs allocation was key to mediating trade-offs across scenarios, highlighting the large potential of spatially explicit approaches for the management of agricultural landscapes.