Global food production is facing increasing uncertainties under climate change and the coronavirus pandemic, provoking challenges and severe concerns to national food security. The role of global agricultural trade in bridging the imbalance between food supply and demand has come to the fore. However, the impact of multifaceted and dynamic factors, such as trade policies, national relations, and epidemics, on the stability of the agricultural trade network (ATN) needs to be better addressed. Quantitatively, this study estimated grouping characteristics and network stability by analyzing the changing global ATN from 1986 to 2018. We found that the evolution of global agricultural trade communities has gone through four stages: the dominance of the US–Asian community, the rise of the European–African community, the formation of tri-pillar communities, and the development of a multipolar community with a more complex structure. Despite witnessing a progressive increase in the nodal stability of the global ATN during the decades, particular gaps can still be found in stability across countries. Specifically, the European community achieved stability of 0.49 and its trade relations were effectively secured. Meanwhile, the remaining leading communities’ stability shows a stable and upward trend, albeit with more significant challenges in trade relations among some of them. Therefore, how to guarantee the stability of trade relations and strengthen the global ATN to resist external shocks has become an essential question to safeguard global food security.

Global agricultural trade plays an essential role in balancing supply and demand regarding agricultural products worldwide. Based on complex network theory, two types of agricultural trade networks weighted by the physical quantity and monetary value were built. In both networks, eight groups of agricultural products showed diverse variation in time and space. During 1986 to 2016, the total physical trade increased by 2.55 times with a gradual growth process, and total monetary value increased 1.98 times with fluctuation. The cumulative distribution of node degree and strength followed power-law distribution. Scale expansion and structure complexity of both networks reflected heterogeneity between nodes and the trend of agricultural economic globalization. Meeting demand and seeking greater returns are the main drivers of global agricultural trade development. Mainly developed countries occupied the important positions in the global agricultural trade network, but some emerging economies such as China, Brazil, and India became important sources of demand and supply. China not only needs to fully use international resources to meet demand for agricultural products, but also needs to ensure its own food security through multiple countermeasures.

The transformation of dietary structure brought about by economic development in populous countries is expected to trigger an increase in grain demand, which will put enormous pressure on the grain supply in these nations and even globally. We simulated nine demand scenarios for 2020–2050 based on China’s dataset for 15 kinds of grains from 1961–2018. The results show that the maximum difference between the predicted grain demand is 323.8 Mt, equal to the total grain consumption of approximately 600 million Chinese people in one year. To capture which demand scenarios will be met when grain productivity gradually improves within reasonable ranges, we present three projections from the production side. In particular, Projection 1 (P1), which maintains productivity at the current level, only fulfills the projected demand for Scenarios 1-LL, 2-LM, 4-ML, and 7-HL and falls short of the maximum value (Scenario 9-HH) by 117 Mt, which requires an additional 250,000 ha of arable land resources to fill the gap. After raising the preset value of grain yield, the productivity of Projection 2 in turn satisfies the demand scenario 5-MM. When both set variables (grain yields and arable area) increase simultaneously, the output of Projection 3 increases by 15.3% over P1. However, it still lags behind the demand of 68 million tons in Scenario 9-HH, thus implying uncertainty in China’s vision of meeting the goal of 95% grain self-sufficiency. Rather than pursuing a single outcome, we discuss multiple possibilities for China’s future grain balance and emphasize the adjusting and compensating role of grain trade and storage in the whole system. Ultimately, this paper calls for a better understanding of the supply–demand gap therein and its future trends to support national grain security as well as global sustainable food policies.

The loess hilly region is a typical ecologically fragile and sensitive area. It is of great significance when measuring the balance between supply and demand of ecosystem services and clarify the relationship between supply and demand of ecosystem services for scientific ecological restoration zoning and realizing regional ecological security and social and economic sustainable development. We take Yuzhong County in the loess hilly region of Longzhong, China as an example; by constructing the matching and coordinating relationship between ecosystem service supply and demand and following the idea of “problem diagnosis—strategic orientation—key optimization”, this paper delimits the ecological restoration zones and proposes corresponding governance strategies. The results show that: (1) The supply and demand of ecosystem services were mainly at a low and medium level, with significant spatial heterogeneity. Overall, the spatial pattern of ecosystem services showed high supply and low demand in the southwest mountainous areas and showed low supply and high demand in the urban gully areas of Yuzhong. (2) There were four types of spatial matching of ecosystem service supply and demand: HH type, LH type, LL type, and HL type. The local spatial autocorrelation of supply and demand was dominated by high-low or low-high spatial misadjustment. The average degree of coordination between supply and demand of ecosystem services was 0.629, indicating that supply and demand were basically coordinated. (3) The ecological restoration zones in Yuzhong County can be divided into six categories: the ecological core protection area, ecological potential restoration area, key ecological restoration area, soil and water conservation restoration area, ecological agriculture development area, and ecological innovation promotion area. The results of this study can provide ideas for the refined management and targeted restoration practice of regional ecosystem services. It provides a strong guarantee for ecosystem management and decision-making in the loess hilly region.

Against the backdrop of intensifying global climate change and increasingly prominent resource constraints, achieving synergies between resilience enhancem