My data for this network visualization comes from Web of Life, a database that holds information relating different species together within symbiotic and parasitic relationships. The specific data I chose from this project revolves around fish and their chosen homes in anemones, tracking which species of fish live most often in the different sea anemones. This data was collected from the waters surrounding Africa, the Indian Ocean, Australia and Tokyo. The bias of this data is simply that it is unclear exactly where the data is taken from, as the listed areas are far from each other as well as expansive. I wish to find a more clear dataset with specifics to better understand the origin of this data and how it was collected. This would allow me to expand this network in an accurate way, pulling in additional species I could be certain lived with the fish and anemones in this dataset. I chose this data to better understand this symbiotic relationship, exploring how diverse the use of anemones are, which are chosen most often and which anemone-dwelling fish are particular in their choices.

Observing the additional visualizations that approach this query, networks in the form of a tree are often used to express this relationship, showing the anemone which holds many different species of fish beneath it. Though this visualization may be more straightforward, it seems to lack the aspect of community that exists in a reef, only allowing for each fish to be associated with one anemone. Similarly, bar charts depicting where a species of fish are most likely to dwell are common, but again this excludes the complexity of the reef and its inhabitants. A network visualization expresses the intertwining of species that exist in these ecosystems.

I began my network with the original dataset, but soon realized I would need to translate the scientific names that this data provided into their more commonly known names in order to make this network more accessible for the public. 

The network here shows the different types of anemone fish, often specifically clownfish, and the anemones that they are found to dwell in. From observing the edges, it’s clear the Leathery Sea Anemone is the most popular among the fish, followed by Merten’s Carpet Anemone and the Bubble Tip Anemone. The Clark’s Anemonefish has the most diverse living spaces while many others only choose one specific anemone to live in.

 I first chose to use the Fruchterman Reingold model for my network, as it gave space for the longer names of species and equally spaced the nodes out. I used the curved edge feature in order to give the feel of a sea anemone blowing in the ocean current and assigned these green. I found it challenging to decide which colors to use for the fish species and anemone species, but settled on orange for the fish, as many of these species are clownfish, and turquoise for the anemones. Though I felt this network was pleasing visually, it did not give much information, as the user is only able to see what certain species choose often for their home, but not able to see which anemones house the most species of fish, or which fish are less particular about their dwelling.

In order to more accurately and clearly show this information, I used the degree in order to size the nodes by how often an anemone is chosen or how many anemones a species will live in. This network makes it much more visible which anemones are more diverse and which fish are more or less picky. Additionally, I changed the model from Fruchterman Reingold to ForceAtlas 2, as this model makes the labels much more legible, while also ensuring minimal overlap for a clearer view. 

In order to improve my network, I would like to use a more interactive software that would allow my nodes to be clickable, where an image of the fish or anemone would be shown. This would give the user a more rounded view of this data, as they would be able to visualize the biodiversity or link their previous knowledge of fish to the data. I also would like to be able to size the labels of the nodes in relation to the degree to emphasize this information. Unfortunately, on Gephi I found it difficult to do this without causing messiness over the network. I would need to use a different software where I could place the labels in a more practical place. I think it would be interesting to find more specific data in order to vary the thickness of the edges by the popularity of anemones, giving thicker edges to the connection between fish and anemone that is most popular. I would also like to continue this exploration and expand it to additional animals and coral living in these communities. A wider network connecting creatures in a different way, possibly using additional colors for edges to communicate food chain or other symbiotic relationships in the ocean would give a fuller picture of how the creatures interact with each other to build this ecosystem.