Fruits important for network cohesiveness may not be the same as those key for bird diets

Understanding how the extinction of mutualistic partners affects species and their functional role in the community is a foundational question in ecology. Such changes impact the architecture of interaction networks, but it remains unclear how they affect interaction outcomes, such as the nutrient resource provisioning for animals. Here, we fitted a community-level frugivory model with a comprehensive dataset of plant-bird interactions in six communities from the Yungas Forest in Argentina. Then, we predicted bird fruit choices before and after plant loss allowing for rewiring and quantified changes in network architecture and interaction outcomes for frugivores (i.e. nutrient composition of diets). Using this modelling approach, we tested whether plants important for network cohesiveness also sustain the nutritional intake of birds. Also, we linked changes in networks and bird diets to the total consumption rates of plants and the nutrient content of their fruits. Finally, we evaluated how the ability of frugivores to exploit new resources after partner loss (i.e. rewire) protected them from nutritional changes in diets. Our results show that plant species that sustain network cohesiveness differ from those that determine the nutritional characteristics of bird diets. Most nutritious fruits and those offering distinctive nutrient combinations within communities were key for the nutrient intake of birds, but they were often overlooked by network metrics. In addition, we show that frugivores that more easily exploit new resources after partner loss can suffer important shifts in the average nutrient content of their diets. Our model suggests that key species for network architecture may not be the same as for bird diets. Thus, we should remain cautious when inferring the functional consequences of plant species loss based solely on their position within networks. In addition, we show that rewiring can affect birds' nutrition, which may carry some costs. Overall, our study reveals mechanistic links between extinctions, network architecture and interaction outcomes, which challenge some assumptions frequently made in frugivory studies.