Background
For months and months, I’ve been seeing and hearing coverage about the Spotted Lanternfly (e.g. See it? Squish it!) on local news and at various landmarks around the city. I saw several of the Lanternfly bug graveyards covering Manhattan streets this summer. The Spotted Lanternfly is so well-known that a costume featuring even it won “Best in Show” this year at the Thompson Square Halloween Dog Festival.
According to New York State Integrated Pest Management, while they are basically harmless to people and animals, “spotted lanternflies are a significant economic and lifestyle pest for residents, businesses, tourism, forestry, and agriculture.” Part of the issue is their exponential growth and swarming behavior. In short, they quickly consume and often then fully destroy certain plants; they are especially damaging to grapevines, among other crops, which is of particular economic concern to New York’s agriculture industry. The public information campaign circulating NYC has revolved around directing New Yorkers to stomp them on sight.
Given the spread of this public messaging and people’s newfound awareness of the Lanternfly over the past year or so, I was interested in mapping Spotted Lanternfly sightings across New York state over time. As I read up on the issue, it turns out that the spread of this particular bug could be directly related to another invasive plant species known as the “Tree of heaven,” one of the Spotted Lanternfly’s preferred plant homes. Through further research, I also began to learn more about myriad other invasive species that conservationists and scientists have been tracking with some worry, due to their destructiveness.
With that, I decided to expand my scope to try and tell an interesting visual story with spatial data about how the top invasive plant and animal species have spread across New York state over the past 10-20 years. While I found some robust recent data concerning the Spotted Lanternfly’s spread, I found far more about other, possibly less well-known, invasive plant and animal species that have been wreaking havoc on natural habitats for decades.
Materials
To work with data on both Spotted Lanternflies as well as other invasive species, I created an account and downloaded datasets from iMapInvasives (https://www.nyimapinvasives.org/data-and-maps), which I found through readings NY state government publications on the Lanternfly. iMapInvasives is “a collaborative GIS-based invasive species database for presence, not-detected, and treatment records” on invasive species nationwide. It is checked and verified by taxonomists and the data is open for use, with certain terms and conditions attached. I downloaded two .csv files: one with all invasive species data bounded by the state of New York, and the other only concerning Spotted Lanternfly sightings. I used Excel to do an initial review of the datasets, then I used ArcGIS to create a Web Map with feature layers using this spatial data, and to then style the layers to create a narrative structure. (Initially, I tried to link to their API to work with their entire dataset, but found myself limited by the formatting built into their Web Map.)
Methods
After I downloaded the data from iMapInvasives and opened it in Excel, I checked that spatial coordinates were included so that the data would appear correctly in ArcGIS. All looked correct at first, with x and y coordinates marked as column headers, however when I uploaded it into ArcGIS as a feature layer, I ended up viewing data in Antartica. That was unexpected (and wrong)! Through some digging, I found that the latitude and longitude were switched in the file, so I overwrote it and recreated my ArcGIS map. Before I kept working on a new Web Map that I titled “NY Invasive Species,” I updated the content description with iMapInvasives’ terms of use and required citations, as noted on their website.
As I began to design the map, I wanted to choose a base layer that was not too busy, but that would show some topographic information so that the viewer would be able to see the basic geography of the state. My thought was that this might provide more context about the distribution of invasive species (e.g. around forests or rivers), but at the same time, I didn’t want to take away from the priority of showing the concentration of species’ observed locations. The basemap I chose features a subdued palette and slightly more topographic detail than a flat map, so that the layers featuring locations of reported invasive species are still the priority when viewed. I then created a series of layers, some of which were helpful to view one at a time, but which could also tell more of a story when layered on top of each other.
For all the layers, I edited Field names in pop-ups to be more readable (e.g. changing “common_name” to “Common Name”) and re-ordered and deleted fields to relate pop-ups to the specific layer being viewed. For instance, in the “Reported Invasive Species by the Number” layer, I put the variable for the number of observations at the top of the list, just under the Common and Scientific Names. I kept the Common and Scientific Names at the top of all of the pop-ups for consistency, and because I judged those to be the highest priority information regardless of the layer’s structure. I also kept other fields across the layers, such as the date of observation and species type (plant or animal). To automatically adjust the pop-up based on the observation point clicked by a user, I added “{Common Name} Sighting” to the pop-up title to call on that variable when clicked.
The layers I created included these, among others:
Reported Invasive Species by Number: Using counts and amounts to visually display large amounts of observed species data, this pulls out which species were most-observed in high quantities while also showing the geographic distribution of the whole dataset across the state. I chose orange for contrast and to suggest a caution, since these are “invasive” species, and I adjusted the size of the circles to make sense for this dataset (some species are observed in the thousands while others are in single digits, and a user should be able to see both).
Top 3 Invasive Species: Filtering for the top three reported invasive species, this shows their relative distribution across New York state. I chose a neutral “earthy” color scheme to match the nature of the subject, and because I was displaying three species in three different colors and didn’t want to overwhelm the eye.
Tree of Heaven Sightings / Spotted Lanternfly Sightings (two layers): Dropping points at observed sightings, together these two layers show specific concentrations of Spotted Lanternfly sightings while relating this to observations of their reported favorite plant home, the Tree of Heaven (another invasive species).
Results & Interpretation
The full map on ArcGIS online can be viewed here (you must belong to the Info-658 Pratt Group).
After taking a look at the various displays across map layers, I was able to see some interesting patterns emerge. There seemed to be some correlation between the Tree of Heaven and Spotted Lanternfly data, but not always, and especially not when zooming out from the New York City area. There was certainly a high amount of reported Lanternfly sightings in the NYC metro area relative to the rest of the state. Could this be due to higher concentrations of the bug, or just greater awareness and a higher population of observers?
By far, invasive plants were observed in much higher quantities across the state than animals. This is perhaps not surprising, especially considering some plants have many pieces to them (e.g. reeds and grasses). Distribution of all the invasive species types seemed to cluster near more populous areas and near bodies of water, but again, this may be the effect of just having more observers located nearby.
Reflection
Generally, I found investigating this dataset, and the community of researchers and scientists maintaining it, compelling. There are so many invasive species being tracked, and some are now so commonplace that I would not have guessed that they weren’t native, or that they were in fact harming other species. (Also, along the way, I did learn that the Spotted Lanternfly is likely less harmful than we originally thought, but more information is needed.)
Methods-wise, I’d like to spend more time experimenting with layers, transparency, and aggregation, with this dataset; these are features in ArcGIS that I tested out but didn’t have as much time to dive into. I think I could create some more interesting polygon visualizations by county or region, for instance, and delegate which layers match with each other versus which should be viewed alone.
I’d also like to dig a little into the iMapInvasives data moving forward to understand more about what is clearly a multistep vetting process for getting these observations published on their map. For instance, for the invasive grasses that are mapped, are they basing the “number observed” on an square footage of grass observed by one person? Is this more or less an estimated data point? I’m also curious to know more about the Spotted Lanternfly data, and if the amount of local knowledge and media coverage had a strong effect on this map in particular. My guess is yes, it had a measurable effect, just from what I’ve observed anecdotally, but it would be interesting to do an analysis of the strength of that effect.