Assessing the impacts of rewilding 

The UK is one of the most nature depleted countries in the world: 60% of the its species have declined over the last 50 years, and many keystone species have been extirpated over the centuries. These losses have had profound knock-on impacts on ecosystem resilience, health, and services.

 

Rewilding is a promising new approach that aims to reintroduce species to restore ecological processes, allowing nature to take its course without human management. It shows great potential in improving biodiversity and ecosystem services, making it an effective, low-cost natural solution for climate change. However, rewilding is a fairly new approach and concerns have been raised about the lack of empirical evidence on its potential outcomes.

 

In collaboration with the Knepp Wildland Project in West Sussex, we are developing ecosystem models to assess the ecosystem wide impacts of rewilding. 

The model

The model is a system of ODEs that uses the Lotka-Volterra equations and a technique called Ensemble Ecosystem modelling. It is informed by quantitative, time-series data from Knepp, as well as qualitative expert knowledge. 

We first back-cast the ecosystem dynamics observed at Knepp before and after species reintroductions, and then make predictions about the range of possible future ecosystem dynamics over a twenty-five year timespan. We aim to answer key questions such as: how has rewilding affected habitat structure, ecosystem services, and ecosystem resilience, and how might it continue to do so in the future? How sensitive is the ecosystem to management choices, such as herbivore stocking rates? How many herbivores are needed to drive vegetation changes, such as a phase shift from scrubland to grassland?

The diagram above shows the core elements of the model. A full description of the model is forthcoming. The model was created by Emily Neil (supervised by Richard Bailey and Ernesto Carrella) and run on Python. It is available on GitHub.

Ongoing work

Model-building and testing is currently underway, and results will be reported in the near future.

© 2021 CoHESyS-Lab