### Abstract

Two predator-prey model formulations are studied: the classical Rosenzweig–MacArthur (RM) model and the Mass Balance (MB) chemostat model. When the growth and loss rate of the predator is much smaller than that of the prey these models are slow–fast systems leading mathematically to a singular perturbation problem. In contradiction to the RM-model, the resource for the prey is modelled explicitly in the MB-model but this comes with additional parameters. These parameter values are chosen such that the two models become easy to compare. In both models the transcritical bifurcation, a threshold above which invasion of predator into prey-only system occurs, and the Hopf bifurcation where the interior equilibrium becomes unstable leading to a stable limit cycle occur. The slow-fast limit cycles are called relaxation oscillations which for increasing differences in time scales leads to the well known degenerated trajectories being concatenations of slow parts of the trajectory and fast parts of the trajectory. In the fast–slow version of the RM-model a canard explosion of the stable limit cycles occurs in the oscillatory region of the parameter space. To our knowledge this type of dynamics has not been observed for the RM-model and not even for more complex ecosystem models. When a bifurcation parameter crosses the Hopf bifurcation point the amplitude of the emerging stable limit cycles increases. However, depending of the perturbation parameter the shape of this limit cycle changes abruptly from one consisting of two concatenated slow and fast episodes with small amplitude of the limit cycle, to a shape with large amplitude of which the shape is similar to the relaxation oscillation, the well known degenerated phase trajectories consisting of four episodes (concatenation of two slow and two fast). The canard explosion point is accurately predicted by using an extended asymptotic expansion technique in the perturbation and bifurcation parameter simultaneously where the small amplitude stable limit cycles exist. The predicted dynamics of the MB-model is in a large part of the parameter space similar to that of the RM-model. However, the fast–slow version of MB-model does not predict a canard explosion phenomenon.

Original language | English |
---|---|

Pages (from-to) | 93-110 |

Number of pages | 18 |

Journal | Mathematical Biosciences |

Volume | 301 |

Early online date | 22 Apr 2018 |

DOIs | |

Publication status | Published - 1 Jul 2018 |

### Fingerprint

### Keywords

- Aggregation methods
- Asymptotic series expansion
- Canard
- Geometrical singular perturbation theory
- Predator-prey models

### Cite this

*Mathematical Biosciences*,

*301*, 93-110. https://doi.org/10.1016/j.mbs.2018.04.006

}

*Mathematical Biosciences*, vol. 301, pp. 93-110. https://doi.org/10.1016/j.mbs.2018.04.006

**Modelling, singular perturbation and bifurcation analyses of bitrophic food chains.** / Kooi, B. W.; Poggiale, J. C.

Research output: Contribution to Journal › Article › Academic › peer-review

TY - JOUR

T1 - Modelling, singular perturbation and bifurcation analyses of bitrophic food chains

AU - Kooi, B. W.

AU - Poggiale, J. C.

PY - 2018/7/1

Y1 - 2018/7/1

N2 - Two predator-prey model formulations are studied: the classical Rosenzweig–MacArthur (RM) model and the Mass Balance (MB) chemostat model. When the growth and loss rate of the predator is much smaller than that of the prey these models are slow–fast systems leading mathematically to a singular perturbation problem. In contradiction to the RM-model, the resource for the prey is modelled explicitly in the MB-model but this comes with additional parameters. These parameter values are chosen such that the two models become easy to compare. In both models the transcritical bifurcation, a threshold above which invasion of predator into prey-only system occurs, and the Hopf bifurcation where the interior equilibrium becomes unstable leading to a stable limit cycle occur. The slow-fast limit cycles are called relaxation oscillations which for increasing differences in time scales leads to the well known degenerated trajectories being concatenations of slow parts of the trajectory and fast parts of the trajectory. In the fast–slow version of the RM-model a canard explosion of the stable limit cycles occurs in the oscillatory region of the parameter space. To our knowledge this type of dynamics has not been observed for the RM-model and not even for more complex ecosystem models. When a bifurcation parameter crosses the Hopf bifurcation point the amplitude of the emerging stable limit cycles increases. However, depending of the perturbation parameter the shape of this limit cycle changes abruptly from one consisting of two concatenated slow and fast episodes with small amplitude of the limit cycle, to a shape with large amplitude of which the shape is similar to the relaxation oscillation, the well known degenerated phase trajectories consisting of four episodes (concatenation of two slow and two fast). The canard explosion point is accurately predicted by using an extended asymptotic expansion technique in the perturbation and bifurcation parameter simultaneously where the small amplitude stable limit cycles exist. The predicted dynamics of the MB-model is in a large part of the parameter space similar to that of the RM-model. However, the fast–slow version of MB-model does not predict a canard explosion phenomenon.

AB - Two predator-prey model formulations are studied: the classical Rosenzweig–MacArthur (RM) model and the Mass Balance (MB) chemostat model. When the growth and loss rate of the predator is much smaller than that of the prey these models are slow–fast systems leading mathematically to a singular perturbation problem. In contradiction to the RM-model, the resource for the prey is modelled explicitly in the MB-model but this comes with additional parameters. These parameter values are chosen such that the two models become easy to compare. In both models the transcritical bifurcation, a threshold above which invasion of predator into prey-only system occurs, and the Hopf bifurcation where the interior equilibrium becomes unstable leading to a stable limit cycle occur. The slow-fast limit cycles are called relaxation oscillations which for increasing differences in time scales leads to the well known degenerated trajectories being concatenations of slow parts of the trajectory and fast parts of the trajectory. In the fast–slow version of the RM-model a canard explosion of the stable limit cycles occurs in the oscillatory region of the parameter space. To our knowledge this type of dynamics has not been observed for the RM-model and not even for more complex ecosystem models. When a bifurcation parameter crosses the Hopf bifurcation point the amplitude of the emerging stable limit cycles increases. However, depending of the perturbation parameter the shape of this limit cycle changes abruptly from one consisting of two concatenated slow and fast episodes with small amplitude of the limit cycle, to a shape with large amplitude of which the shape is similar to the relaxation oscillation, the well known degenerated phase trajectories consisting of four episodes (concatenation of two slow and two fast). The canard explosion point is accurately predicted by using an extended asymptotic expansion technique in the perturbation and bifurcation parameter simultaneously where the small amplitude stable limit cycles exist. The predicted dynamics of the MB-model is in a large part of the parameter space similar to that of the RM-model. However, the fast–slow version of MB-model does not predict a canard explosion phenomenon.

KW - Aggregation methods

KW - Asymptotic series expansion

KW - Canard

KW - Geometrical singular perturbation theory

KW - Predator-prey models

UR - http://www.scopus.com/inward/record.url?scp=85046153668&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85046153668&partnerID=8YFLogxK

U2 - 10.1016/j.mbs.2018.04.006

DO - 10.1016/j.mbs.2018.04.006

M3 - Article

VL - 301

SP - 93

EP - 110

JO - Mathematical Biosciences

JF - Mathematical Biosciences

SN - 0025-5564

ER -