Habitat fragmentation is modifying landscapes and the distribution of floral resources, possibly shaping also pollinator resource acquisition. Here, using urban parks as field laboratories for the dramatic contrast around, we aimed to clarify how fragmentation and local flower availability shape bumblebee foraging dynamics by characterizing several components: the nutritional content and plant composition of collected pollen pellets, the foraging rate and the plant-nutrition association along a fragmentation gradient. We found mostly negative linear or non-linear relationships between nutritional quality and fragmentation, tight plant composition-nutrition associations interpretable as low access to alternative resources, and shorter foraging time in smaller green areas, showing behavioral limits by the landscape. Thus, fragmentation can constrain all aspects of bumblebee foraging by compromising resource accessibility. This study illuminates the link between landscape features and the nutritional ecology of pollinators, a key aspect for understanding pollinator foraging dynamics and even for outlining mitigation measures in urban contexts.

Habitat fragmentation is known to affect biodiversity, but the impact on pollinators and their interactions with plants is still unclear in anthropized landscapes. Islands are open-air laboratories for ecological studies with simplified communities and interactions, suitable to disentangle how land-use alteration impacts pollination ecology and its ecosystem service. Here, we used Maldives islands as model systems to investigate how pollinator richness, their mutualistic interactions with plants, and pollination efficiency are shaped by the degree of green area fragmentation (i.e., gardens, parks and semi-natural green covered patches), by considering both community- and species-level responses. To do this, we surveyed pollinators from 11 islands showing a gradient of green area fragmentation. In order to characterize the interactions between plants and pollinators and obtain a novel and comprehensive view of the key ecological dynamics, a DNA metabarcoding approach was adopted to identify the pollen carried by pollinators. We found that green area fragmentation at intermediate levels played positive effects on pollinator richness. However, fragmentation decreased interaction network complexity. Intriguingly, body size mediated the effect of landscape alteration on plant-pollinator interactions, as only the largest bee species expanded the foraging breath in terms of transported pollen richness at increasing fragmentation. In parallel, the pollination efficiency increased with pollinator species richness in two sentinel plants. This study shows that moderate landscape fragmentation of green areas shapes the ecosystem service of pollination, where in spite of interactions being less complex and mediated by pollinator body size, pollinator biodiversity and potential plant reproduction are supported.

As human activities on our planet persist, causing widespread and irreversible environmental degradation, the need to biomonitor ecosystems has never been more pressing. These circumstances have required a renewal in monitoring techniques, encouraged by necessity to develop more rapid and accurate tools which will support timely observations of ecosystem structure and function. The World Exposition (from now ‘EXPO2015’) hosted in Milan from May to October 2015 was a global event that could be categorized as a mega-event, which can be defined as an acute environmental stressor, possibly generating biodiversity alteration and disturbance. During the six months of EXPO2015, exhibitors from more than 135 countries and 22 million visitors insisted on a 1.1 million square meters area. Faced with such a massive event, we explore the potential of DNA metabarcoding using three molecular markers to improve the understanding of anthropogenic impacts in the area, both considering air and water monitoring. Furthermore, we explore the effectiveness of the taxonomy assignment phase considering different taxonomic levels of analysis and the use of data mining approaches to predict sample origin. Unless the degree of taxa identification still remains open, our results showed that DNA metabarcoding is a powerful genomic-based tool to monitor biodiversity at the microscale, allowing us to capture exact fingerprints of specific event sites and to explore in a comprehensive manner the eukaryotic community alteration. With this work, we aim to disentangle and overcome the crucial issues related to the generalization of DNA metabarcoding in order to support future applications.