Undergrad Project
During my undergrad studies, I joined the Ecological Complexity Lab headed by Dr. Shai Pilosof, a group dedicated to the study of ecological networks. My project focused on species roles in a network of hosts and parasites, and the ways these roles change over time.
​
Background
​
Ecological networks are a useful way to describe the many interactions between species in a community, whether we are interested in trophic interactions (e.g. food webs), mutualism (e.g. pollination), or host-parasite interactions. These networks often have a modular structure, with groups of species that are highly connected to each other and less connected to species outside the group.
The role a species plays in a network can be defined by its connectivity within its own module and across modules. Exploring species' roles may help us identify key species in the community.
​
Previous studies of the modularity and species' roles of ecological networks have focused on either snapshots of disconnected time points, or on aggregated networks (treating the network as if all interactions were occurring simultaneously). These approaches might lead to wrong conclusions about community dynamics. A possible way to address this issue is using multilayer temporal networks, in which the layers represent the same community at different time points.
In my project, I analyzed a network of small mammals and their ectoparasites, based on data collected over 6 years in western Siberia.
Created with biorender.com
An example of multilayer network structure. Each layer represents a year, and the edges (full lines) within a layer represent the interactions that were recorded in that year. Interlayer edges (dashed arrows) connect each species in a given layer to itself in the next layer, if it occurs there.
​
I found that out of 74 species occurring in multiple years, 48% changed their roles at least once. In contrast, only 23% moved between modules.
​
This analysis indicates that species roles indeed change over time, but no clear patterns of change were observed. The network is fragmented into many small modules, which likely explains the high number of connector species.
​
I also identified seven network hubs. They were all hosts – six rodents (families Muridae and Cricetidae) and one insectivore (family Talpidae). Each had this role in one layer only. One possible conclusion might be that being a hub depends on environmental factors at a given time, and less on some inherent characteristic of the species.
​
See the project poster here.
​
Species roles over time: Each rectangle represents a species. Species below the orange line are hosts; the rest are parasites.
Peripherals (red) have few links in general; connectors (blue) have few links within their modules but connect them to other modules; and network hubs (green) are highly connected both inside and outside their modules.