Crop nitrogen (N) demand is defined by various physiological, management and environmental factors that interact with each other and involve temporal and spatial dynamics. Variable rate N application (VRNA) intends to address this but the underlying algorithms often remain rather rigid and deterministic, are partly subject to high uncertainties and largely leave the agronomic expert’s knowledge and experience unnoticed. A novel generic system architecture was conceptualized to overcome these limitations and respond to the complexity of N management in a straightforward and both, reactive and proactive manner. Having a fuzzy expert system as methodical core, the approach mainly relies on human input to grasp the circumstances at a specific N application and address the required parameter interactions. As a holistic concept, it further aims at a high versatility in terms of considered input data and utilization with different sensor and application technology, as well as a digitized VRNA process chain involving graphical user interfaces to simplify procedures for data presentation, decision making, application and documentation. Bringing the presented concept into a prototypic implementation considering real-time crop N sensor and mapped soil data, its consistency was verified. At the same time, potential functionalities, as well as limitations were illustrated and technical requirements on specific subsystems were clarified by the prototype. The possible risks that stand in the way of high benefits due to a strong focus on expert knowledge can be countered by using digital tools and heading towards hybridization with other VRNA approaches.

The post-2020 global biodiversity framework calls for a transformative change in food systems. Promoting agricultural multifunctionality is a viable approach to this sustainability transformation. The eastern Qinghai–Tibetan Plateau (QTP) is both one of the world’s largest livestock grazing systems and a hotspot of endemic birds in Asia. In this research, we aim to investigate the impact of livestock grazing on alpine bird assemblages at the local scale (alpha diversity) and their variation across the pastoral landscape (beta diversity). In the study area Nyanpo Yutse, we conducted surveys of 126 bird sample plots during two breeding seasons to acquire bird assemblage data. Meanwhile, we employed unmanned aerial vehicles to measure 2D and 3D habitat features within the 150-m radius. We investigated the key habitat variables driving the spatial distributions of both alpha and beta diversities of birds. Particularly, we partitioned beta diversity into its turnover and nestedness components and tested their patterns across sites of four levels of livestock grazing intensities (LGIs). Our results found no significant correlation between LGIs with species richness of birds, while 2D and 3D habitat complexity and built structure were positively correlated with alpha diversity (p < 0.05). At the landscape scale, pairwise LGI differences had no significant correlation (p > 0.05) with any pairwise beta diversity. The ordination plotting detected distinguished habitat preferences among 12 common birds and eight endemic birds. The multiple-site beta diversity of the 126 plots showed high species turnover (>0.871) where LGI was lower than 1.065 sheep units/ha, indicating the importance of moderate grazing for the conservation of diverse avian assemblages at the landscape scale. Our study demonstrated that extensive pastoralism is important for both maintaining the mosaic landscape and conserving avian biodiversity on the eastern QTP. We unveiled one of the ecological mechanisms through which synergies can be realized to support both agricultural production and biodiversity conservation in the Tibetan grazing system.

Aside from classical meteorological measurements from a wide variety of weather stations, so-called crowdsourcing approaches have been applied to the collection of meteorological data, as well as generally atmospheric measurements in recent years. The topic of this chapter is crowdsourcing, but it is also a broad overview of sensors applied in crowdsourcing and corresponding projects. This chapter first introduces the definition and typology of the relevant terminology, along with a short history of crowdsourcing. We describe and discuss aspects of both the typically employed low-cost measurement technology, as well as human factors that emerge from laypersons carrying out the measurements. This includes general considerations concerning measurement uncertainty, as well as possible measures to improve the data quality in mobile nonexpert sensing. Devices and sensors, ranging from dedicated devices over do-it-yourself sensors to internal smartphone sensors are discussed, as are a number of applications and projects that illustrate in which ways crowdsourcing approaches have been applied to the measurement of atmospheric parameters and what type of challenges have been addressed and which ones remain.