1

The Pirate Kitty Theory: How a House Cat Being Let Out Led to the Extinction of the Dodo Bird

Dr. David Summers1 , Booyah2

1 Department of Cat Ecology, Cranberry-Lemon University, Pittsburgh, PA, USA

2 Indian Ocean Ecological Department, Greater Somali-East Indies Academic Society of Pirates and Scallywags, Mozambique and Somalia

Abstract

It may not have been the Dutch settlements, or the 17th-century sailors, but one of the sailor’s cats who killed the dodo. The dodo is a flightless bird universally used as a symbol for extinct animals. It lived undisturbed on the remote island of Mauritius and looked like it couldn’t survive modern life outside of a protected remote island. Because of the Dutch settlements, humans have been largely to blame for its disappearance. However, new bone fragment evidence confirmed by predator-prey models, Epic Battle Simulator, and a house cat-trained Generative Adversarial Network[1] show that a single pirate’s cat may have decimated the dodo population beyond recovery in the late 17th century. This wild claim is additionally backed by the tales and exploits from the journal of the infamous Captain Cooksley “Poopdeck” Johnson, shown in Appendices A-F, which track Poopdeck’s life story from a well-spoken English aristocrat to a swashbuckling adventurer filled with more rum than blood. With all this evidence…somehow...we still couldn’t get published in Nature.

Keywords: Dodo, Ecology, Predator-Prey Models, Epic Battle Simulator, Generative Adversarial Networks, Differential Equations, Pirates, Murderous House Cats, Lord Whiskers, Captain Cooksley “Poopdeck” Johnson, Animal Extinction

1. Introduction

In 1598, European sailors first encountered the dodo bird in Mauritius. In under a century, they became extinct. While their extinction was originally blamed on human hunting[2], and then on the destruction of the dodo’s habitat by human agriculture[3], new evidence shows that the lion’s share (or cat’s share) of the blame should be directed toward one cat, named Lord Whiskers, who arrived on The Silver Fox, an English privateering vessel, and flourished in Mauritius.

Late last year, a journal was discovered belonging to the English Pirate Captain Cooksley “Poopdeck” Johnson. Key excerpts of this journal are shown in Appendices A-F. In the journal, he detailed multiple brushes with Dutch merchants and sailors while harboring in hidden coves around the island of Mauritius, narrowly escaping capture while plundering vessels across the Indian Ocean. According to Poopdeck’s account of his life, they brought aboard a vicious cat[Appendix A] who did his own pillaging and raiding while docked on shore on the island.

While the fantastical recounting of Lord Whiskers’ hunting ventures may have not been taken seriously, further analysis shows that what the Dutch did to the dodo’s environment is insignificant in comparison to the savage and brutal escapades of one kitty cat.

2. Poopdeck’s account and other evidence

As shown in the excerpts from Poopdeck’s memoir[4] in Appendices A-F, Captain Cooksley “Poopdeck” Johnson sailed the Indian Ocean in his ship The Silver Fox while it was primarily controlled by the Dutch in the 17th century. With the early trade routes being established by the Dutch East-India Company, there were endless opportunities. After carrying out plundering and general pirating during the day, Poopdeck would often lay low along the coasts of Mauritius, east of Madagascar.

Figure 1.1: Captain Cooksley “Poopdeck” Johnson and his ship The Silver Fox

According to the accounts detailed in his memoir[4], Poopdeck would change up the location of his ports to evade capture. He was unlikely to ever break land in the same location more than once except to retrieve buried supplies. While docked, The Silver Fox’s cat, Lord Whiskers, would be released ashore along with the crew. As shown in the excerpts outlined in Appendices C, D, and E, Lord Whiskers would murder on average one dozen dodos an hour, while the crew would eat less than 10% of the dodo carcasses brought back to the The Silver Fox for the cat’s validation.

3. Predator-prey modeling

This paper will compare and contrast the effects of the Dutch hunters, land use, as well as Lord Whiskers’ personal blood sport through Lotka-Volterra (L-V) modeling as well as advanced simulations and GAN machine learning. For each of the models, an initial population of 37,000 dodo birds was assumed according to the estimate in a study[5] carried out a couple of years ago.

3.1. Lotka-Volterra equations

The classic L-V predator-prey model, the data extracted from the Untitled Memoir[4], and the estimates of the 17th-century dodo population can be easily transformed into the following differential equations, where x is defined by the number of dodos and y is the number of Lord Whiskers (that is, 1):

According to the records shown in Appendix D, Lord Whiskers was allegedly “as old as the ship herself,” which would give us y=1 and = 0.

was then selected to be 37,000 as estimated[5] while was estimated by Poopdeck’s detailed accounting in his journal[4]. With these equations simplified, we decided against using any advanced Ordinary Differential Equations (ODE) models because we were only measuring the rate at which Whiskers killed the dodos, which thankfully wasn’t going to need any Laplace transforms.

3.2. Simulation design and execution

According to the excerpt in Appendix E, when a dodo began to cry out for help while being savagely maimed by Lord Whiskers, all of the nearby dodos would begin racing to its rescue only to also become victims.

Because Whiskers was known to play with his food, this was a regular occurrence at each port in Mauritius. This hunting mechanism allowed for our research team to run accurate simulations using a verified and validated gaming engine and combat simulator, Epic Battle Simulator[6].

As demonstrated by classic simulations such as One Super Sherman Tank vs 2,000,000 Romans[7], 1,000 T-Rex vs 80,000 Chickens[8], and 50 John Wicks vs 2 Million Zombies[9], Epic Battle Simulator was the perfect way to simulate a small number of super soldiers against a large number of powerless units and determine how much time it would take to decimate an entire population of flightless birds.

3.3. Bandit-dodo adversarial network

We decided to attempt to create a Generative Adversarial Network (GAN) model because we kept hearing people talk about it. Since it had the word “adversary” in the name, it seemed perfect for modeling an actual adversary, and it would be nice to get some machine learning into this paper for the buzzwords and academic cred. Unfortunately, we soon realized it didn’t have much to do with the type of adversary we were thinking about, but we were too embarrassed to back down, so we created one anyway. We should have learned our lesson from when we tried to model deforestation in developing countries using a Random Forest algorithm[10].

Thankfully, we had the two preceding models and some data from the vicious cat Bandit, who killed birds at a rapid pace, from the drone defense program in another research study[1] (shown in Figure 1.2).

Figure 1.2: Bandit the cat

We input the results into some PyTorch GAN library and prayed that we would be able to interpret the results.

4. Results

The results from the three separate models are shown in Figure 1.3 where the y axis is the estimated dodo population and t is unlabeled time. The blue line represents the L-V equation model estimate, while the green line represents the results from our Epic Battle Simulator, and the red results show what we think the GAN results are. The uncertainties of each model were graphed as dotted lines but were too large to be seen once we zoomed in on the results.

Figure 1.3: Predator-prey model results

Naturally, the L-V equations created the most trustworthy-looking line due to their simplicity. Unfortunately for the Epic Battle Simulator results, the software was unable to simulate the long naps Lord Whiskers would take on the high seas in between the mass slaughtering of dodos, and only estimated how long it would theoretically take Whiskers to kill all of the dodos on the island without any breaks. Unfortunately, we either didn’t correctly interpret the GAN model or it didn’t work. But since we are not certain, we’re going to blame it on not having enough data as is standard industry practice.

Many skeptics and reviewers of this paper have raised concerns that an adult dodo is too large for a typical cat to hunt and that our simulations are dubious, as well as our trust in Poopdeck’s testimony. To this we answer, Lord Whiskers was no ordinary cat, and they clearly don’t understand how cats work. A murderous cat is the only explanation for the many dodo skeletons found completely intact and in mass graves near our best interpretations of Poopdeck’s buried treasure map locations,...