Last time, we walked through creating an environment for which to simulate piping plovers. We digitized Sauble Beach, made it into a bunch of 1m X 1m cells based on habitat, and populated these cells with some information. If you’d like a refresher on all this, check out last post! Alternatively, here’s a summary graphic of what we explored in the previous post.
Today, we’ll explore how the piping plovers are simulated in the model, and some of the initial experimentation that was done.
Let’s start with the nests. At the beginning of a simulation, the model randomly places a certain amount of nests throughout the given environment. There are, however, some rules:
- The number of nests = half of the total initial adults
- The total initial adults are randomly chosen by the model at the beginning
- Nests can ONLY be placed in an “open beach” habitat type
- Once a nest is placed, no nests can be placed within a 200m radius of it
- This simulates the “territory” that nesting adults create with their nests
These nests, for now, are essentially symbolic and are only used to round up the piping plover chicks at the end of a simulated day for them to sleep.
Once the nests are placed, the piping plovers hatch! Of course, in real life, there would be a period of egg laying and rearing, but we’re skipping that part for now. One of the main metrics I’m tracking is the mass of the piping plover chicks, and I want to introduce stressors that may or may not affect their mass and growth rate. So, when the piping plovers first hatch, the model assigns an initial mass to them. These initial masses are randomly distributed; all the initial masses aren’t the same, but they are usually around 5 grams.
In order for the piping plover chicks to gain energy, they must forage. As I mentioned before, each habitat type has a different foraging quality associated with it.
So, in order to gain the most amount of energy possible (which is crucial for the young of any species), the piping plovers in the model have some “intelligent movement”; that is, they will try to move around their environment to try to get to an enviro-agent with better foraging quality. For example, if a piping plover chick was in an enviro-agent with the habitat type “open beach”, they will base their movement on trying to get to enviro-agents with foraging quality better than “open beach”, such as “intertidal zone”, “stream”, or “dunes”.
To mix things up, we can introduce some humans into the mix. Essentially, the “human presence” is just a number between 0 and 100 we can pass to the model. This corresponds with the “% chance” a human would be present in a given enviro-agent. For example, if we sent the model a 40, that means that each enviro-agent has a 40% chance of a human showing up.
In a real life environment, more humans means that piping plover chicks will spend more time paying attention to and fleeing from humans rather than spending time foraging, presumably resulting in a lower growth rate. In this model, there are two rules for when piping plovers are foraging with humans around:
- If a piping plover chick is in the same enviro-agent as a human, the piping plover chick will not forage; instead, it will flee to a new enviro-agent
- If there are humans present within a 25m of a foraging piping plover, but not in the same enviro-agent, the piping plovers will forage but at a reduced rate.
With all these rules in play, let’s sum it up in a simple flow diagram:
This is the flow diagram that is used for each enviro-agent. To sum up quickly:
- If there are no piping plovers in the enviro-agent, move onto the next one
- Otherwise, if there are piping plover chicks, AND there are humans in the same enviro-agent, move the piping plover chicks to a different enviro-agent without any energy gain
- If there aren’t any humans in the enviro-agent, but there are some nearby, piping plover chicks gain energy at a reduced rate. Otherwise, they gain all energy available to them for that time period
- If it’s time for the piping plovers to rest and they are not at their nest, they begin to move toward the nest. Otherwise, they rest.
These rules happen for each enviro-agent per time step. Time-wise, we are only simulating 31 days for now, but we’ve separated this into 5-minute time steps. So, the model is updated every 5 minutes using the rules above.
So from here, perhaps you can see how we might be able to do some experimentation using our model. The one thing we wanted to look at to get an accurate start is how well the model can predict piping plover growth rate without humans present. To do this, we aimed for the growth rate as calculated at Chaplin Lake, Saskatchewan. Chaplin Lake is a designated Important Bird Area, is used by piping plovers for breeding, and does not see nearly as much human presence. The results of our “calibration” is as follows:
The red line is the growth rate that was observed in Chaplin Lake. This is what we roughly wanted to emulate. And when I say roughly, I really do mean it. The growth rate in Chaplin Lake was still only experimental evidence; it is not necessarily the true growth rate of piping plovers. Nonetheless, it at least provides a good base to experiment against. In this case, we got our model reasonably close to that growth rate as seen in the black line.
As mentioned above, we were able to pass the model a number that corresponds to different “levels” of human presence: a 20 would correspond to a 20% chance of a human occurring in a given enviro-agent, a 40 corresponds to 40% chance, and 60 corresponds to 60%. So, our first experiment was to test how these levels of human presence affected the growth rate. 20 simulations were run per level, and the results are as follows:
As you can see, compared to the base model, as you add more humans to the mix, the growth rate of the piping plover chicks tends to drop. The piping plover chicks are spending more time fleeing from humans than they are foraging.
A common technique to mitigate this is in real life is to set up nest exclosures. These are fences around the nest where humans are not supposed to go. So, we experimented with making a rule that humans could not go within 100m of the nest. Theoretically, this would allow piping plover chicks a “buffer space” where they can forage without the disturbance of humans. Here are the results:
There are still some drops in growth rate as you add more humans, but not nearly as dramatic of a drop. This is easily seen in the following plot:
In this plot, we are only looking at piping plover chick weights at day 31. The black line is the day 31 chick weights with an exclosure, and the red line is the day 31 chick weights without the exclosure. As you move across the anthropogenic level (human presence), both show a decrease in day 31 chick weight, but you can see a more dramatic decrease in weights with no exclosure versus those with exclosures.
So what’s next? Many things! As you can tell, the work we’ve done here is very basic and preliminary. We absolutely need to add more detail in about the environment, piping plover chicks, and piping plover adults. Additionally, we’ll be looking to add in some interspecies interactions (i.e. how do piping plovers deal with other species?), and we’ll be adding in predation. Predation is especially imporant to look at during the egg rearing and hatchling stage. Humans tend to increase “human-subsidized predators”, or predators that increase in numbers in the area as humans increase in numbers in the area (such as gulls, raccoons, weasels, etc.). So, it would be worth investigating and simulating some of those effects as well.
That about wraps up the piping plover simulation studies for now! There are plenty of exciting things that are in the works for 2018, so I’ll be sure to make some new posts as those happen.