Abstract
Ecologists have been compiling ecological networks for over a century, detailing the interactions between species in a variety of ecosystems. To this end, they have built networks for mutualistic (e.g., pollination, seed dispersal) as well as antagonistic (e.g., herbivory, parasitism) interactions. The type of interaction being represented is believed to be reflected in the structure of the network, which would differ substantially between mutualistic and antagonistic networks. Here, we put this notion to the test by attempting to determine the type of interaction represented in a network based solely on its structure. We find that, although it is easy to separate different kinds of nonecological networks, ecological networks display much structural variation, making it difficult to distinguish between mutualistic and antagonistic interactions. We therefore frame the problem as a challenge for the community of scientists interested in computational biology and machine learning. We discuss the features a good solution to this problem should possess and the obstacles that need to be overcome to achieve this goal.
Highlights
Since the early days of the field, ecologists found themselves detailing the multitude of interactions occurring between different populations, such as predation, parasitism, and herbivory or pollination, seed dispersal, and symbiosis ("mutualistic" interactions) [1, 2]
When we project onto this space all the remaining nonecological networks, we find that, despite being "out of fit", the new networks fall exactly where they should be: the method can be used to classify new networks based on their position in the plane
When we compute the three metrics illustrated above for each randomization and project the resulting values on the principal component analysis (PCA) space, we find that almost all networks cluster together under the Erdős– Renyi model (Fig 1); note that ideally these should collapse to a point—the remaining scatter indicates that our approximation is not accurate for very sparse or very small networks, possibly because the size of the fluctuations of these quantities around their mean is influenced by size and connectance
Summary
Since the early days of the field, ecologists found themselves detailing the multitude of interactions occurring between different populations, such as predation, parasitism, and herbivory (all "antagonistic" interactions) or pollination, seed dispersal, and symbiosis ("mutualistic" interactions) [1, 2]. To make sense of these data, they built networks in which nodes are species and edges stand for interactions between species. Whereas the earliest published food web dates back more than a century [3], the past 20 years have seen a sharp increase in both the number of networks published and their quality (i.e., higher level of detail, larger number of species reported, and larger number of interactions, often including weighted edges). Reducing a multidimensional object such as a network to a few numbers is a daunting task, but paralleling the progress of network analysis in other branches of science, ecologists set out to compute summary statistics on the empirical networks they collected; these metrics range from very simple measures, such as the size (number of nodes) and connectance (proportion of realized connections) of the network, to large-scale properties such as modularity As in other branches of science, ecologists investigated degree distributions, motif profiles [8], k-cores [9], and many other network properties
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
