1 of 37

BI 559 Lecture 10: Multi-species Pt. II

Protoss

Zerg

resources

Starcraft 2 (?)

Today:

  • Look at all the stable states for a 2-species system predicted by the Lotka-Volterra model�
  • Go through the results of a paper that tests these predictions in a changing experimental environment
    • How does competition change when there is an external pressure that inhibits all cells?

2 of 37

From last time

 

 

1

2

 

 

 

 

Lotka-Volterra model

Growth rate of each species’ population

Per-capita growth rate: how fast each species doubles

Carrying capacity for each species: its maximum density under the current nutrient conditions

Competition coefficients: how strongly each species inhibits the other

3 of 37

 

 

 

 

There is a stable equilibtrium with non-zero densities of species 1 and species 2!

“coexistence”

Stable points: where the growth rates of each species are 0!

4 of 37

Coexisting species

Phase plot

Time-axis growth curve solutions

5 of 37

 

 

Model:

Phase plane:

 

 

Data:

When do we get stable coexistence?

6 of 37

 

 

Need two conditions:

 

and

 

 

 

 

If mutual inhibition is “low”, we get coexistence!

Let’s look at the other possibilities!

7 of 37

One species dominates

Phase plot

Time-axis growth curve solutions

When does one species dominate?

8 of 37

 

 

Need two conditions:

 

and

 

 

 

 

 

9 of 37

Third scenario:

 

 

 

 

Species 1 can dominate or species 2 can dominate. It depends on the initial cell densities!

“bistability”

10 of 37

 

 

 

 

Need two conditions:

 

and

 

 

If both species are competitive we get bistability!

11 of 37

To summarize

coexistence

bistability

single-species dominance

 

 

 

12 of 37

Is Lotka-Volterra a good model for 2-species competition and growth?

  •  

13 of 37

Today’s paper

Can the Lotka-Volterra model allow us to understand multi-species growth under changing conditions?

Can the Lotka-Volterra model predict multiple phases of bacterial growth?

14 of 37

Abreu et al.’s question:

If species compete

 

 

1

2

 

 

but there is also an external agent killing both,

External agent

how does that effect who dominates or if they coexist?

vs

?

15 of 37

Why ask this question?

Species competing in a gut environment:

broad-spectrum antibiotic

Neither species is able to develop high density as it normally would. How does this affect mutual inhibition?

16 of 37

Why ask this question?

Species competing in an ocean environment:

Changing ocean temperature

How can we add such a global mortality effect into the LV model?

What does the model predict?

17 of 37

Abreu et al.’s hypothesis

 

 

Low competition → coexistence

Both species competitive → bistability

 

 

One species competitive → dominance

 

They hypothesize that in high-mortality conditions (e.g. antibiotics), fast growers will tend to dominate more!

How do you incorporate this into the model and then test it experimentally? We’ll step through both.

18 of 37

How do you use the Lotka-Volterra model to interrogate this question?

 

 

1

2

 

 

 

 

 

 

 

19 of 37

A little math to get the equations into the Abreu et al. form:

 

 

 

 

 

 

(you’ll see why they did this on the next slide)

 

 

 

 

 

20 of 37

Why normalize the population density variable?

It makes for a very simple interpretation of the competition coefficients:

 

 

 

 

 

 

Stable coexistence if:

 

 

1

1

 

 

Now stable coexistence if:

 

 

21 of 37

Abreu et al.’s new stability conditions:

coexistence

bistability

single-species dominance

 

 

 

22 of 37

Abreu et al. model to investigate competition in a harsh environment

Now they do something very clever. They rearrange their equations (this is a lot of math) to define new competition parameters that aren’t constants anymore. . . .

23 of 37

Abreu et al. model to investigate competition in a harsh environment

In this new model, competition is not a constant, but depends on:

  • Species growth rates
    • Experimentally manipulated using different strains!
  • Global mortality rates
    • Experimentally controlled via dilution rate (we’ll see how)!

24 of 37

Abreu et al. model to investigate competition in a harsh environment

 

25 of 37

Abreu et al.’s experiment

5 soil bacteria species:

Pseudomonas putida (Pp)

(sciencephoto.com)

Enterobacter aerogenes (Ea)

(microbewiki)

Pseudomonas aurantiaca (Pa)

(Samina Mehnaz (2013))

Pseudomonas citronellolis (Pc)

(credit Dennis Kunkel)

Pseudomonas veronii (Pv)

(Montes, et al. (2016))

26 of 37

Abreu et al.’s experiment

5 soil bacteria species:

(couldn’t find PC colony image)

Distinct colony morphologies allow plating of mixed cultures and counting colony-forming units of each species.

27 of 37

Abreu et al.’s experiment

Grown in a defined medium:

  • Glucose
  • NH4
  • Cholesterol
  • K2HPO4
  • KH2PO4
  • K-citrate
  • CaCl2
  • MgSO4
  • Trace metals (iron, manganese,etc)

 

 

 

Gives a wide range of per-capita growth rates to probe the predictions of their LV model!

28 of 37

Abreu et al.’s experiment

E. aerogenes

P. veronii

 

 

 

 

29 of 37

Implement mortality by diluting culture

E. aerogenes

P. veronii

time

At regular time intervals, remove a small fraction of the mixed culture and dilute into fresh medium and discard the rest.

time

 

30 of 37

Abreu et al. model to investigate competition in a harsh environment

Measured for each strain on its own

Measured in constant condition experiments

Controlled experimentally

Controlled or measured!

31 of 37

What does the model predict?

 

 

 

Can experimentally control this.

Can measure this!

32 of 37

A very elegant way of plotting this prediction

 

 

 

Bistability

Winner depends on initial fraction

Coexistence

Fast wins

Slow wins

Each point on this plot represents an entire 2-species system.

Its position corresponds to what stable point the L-V model predicts.

A phase diagram of phase diagrams!

 

 

 

 

33 of 37

The model predicts specific transitions

These two conditions can be realized with their strain combinations.

This can be manipulated by increasing the dilution (i.e. death/mortality) rate!

Then they can measure whether they observe these transitions.

 

Each point on this plot represents an entire 2-species system.

Its position corresponds to what stable point the L-V model predicts.

A phase diagram of phase diagrams!

34 of 37

What do they observe experimentally?

35 of 37

What do they observe experimentally?

36 of 37

How did Lotka-Volterra do?

  1. Did we observe the 3 stable points?�����
  2. Did we observe the proper transitions between them?

One species wins:

Coexistence:

Bistability:

37 of 37

What have we learned?

coexistence

bistability

single-species dominance

The Lotka-Volterra model predicts three kinds of 2-species systems

With the correct extension to the model and a clever experimental system, we can understand the complex dynamics of species competing in environments with harsh mortality!

In environments with high mortality, it is better to be a fast grower!

It’s a simple model, but maybe not bad!