HEAD OF COMPLEX SYSTEMS GROUP
It is the last lesson of modern science, that the highest simplicity of structure is produced, not by few elements, but by the highest complexity.
—Ralph Waldo Emerson
Goethe; or, the Writer (1850)
APPLICATIONS
- Precision livestock production: Prey-predator Grassland-Livestock (PPGL) model
Precision livestock production (PLP) or the manipulation of livestock activity taking into account the different components of agroecosystems -plant, animals and plant-animal interactions- to improve production has acquired growing importance in recent years within an “ecological intensification” new paradigm. In particular, there has been an increasing interest in applying mathematical modeling to support PLP.
We are developing an integral ecological approach to PLP by modeling the dynamics of the combined grass-animals system as a prey-predator dynamical system. Our prey-predator grass-livestock system PPGL model involves two variables, the grass height and the individual live weight of animals, as well as the nonlinear interaction between them: animal performance (liveweight variations & reproduction) is linked with grass consumption, which depends on forage availability, which in turn is affected by the grazing pressure.
- Overyielding of artificial grasslands: Generalized Interactions Lotka-Volterra (GILV) model
-
We modelled communities of perennial crop mixtures by using a generalized Lotka-Volterra model,
i.e. a model such that the interspecific interactions are more general than purely competitive. We
estimated model parameters -carrying capacities and interaction coefficients- from, respectively, the
observed biomass of monocultures and bicultures measured in a large diversity experiment of seven
perennial forage species in Iowa, United States. ​
2. The sign and absolute value of the interaction coefficients showed that the biological interactions
between species pairs included amensalism, competition, and parasitism (asymmetric positive-negative
interaction), with various degrees of intensity.
3. We tested the model fit by simulating the combinations of more than two species and comparing
them with the polycultures experimental data. Overall, theoretical predictions are in good agreement
with the experiments.
4. Using this model, we also simulated species combinations that were not sown. From all possible
mixtures (sown and not sown) we identified which are the most productive species combinations.
5. Our results demonstrate that a combination of experiments and modelling offer a scientific method for
designing overyielding and sustainable grassland mixtures.