Research

Metapopulation ecology and management

Anthropogenic (but also natural) processes often lead to the fragmentation of habitats and populations, the latter being termed metapopulations. The habitat fragments may further be subject to dynamics (degradation, human and natural disturbance, turnover) that affect the population dynamics. Main questions of interest are the survival of metapopulations in static and dynamic fragmented landscapes, and how survival is affected by different land-use and conservation strategies. Metapopulation models often serve as ecological target functions in an ecological-economic modelling context.

Economic modelling of conservation policies

Market-based instruments such as compensation payments or tradable land-use permits are gaining increasing popularity in research and practice. To assess the suitability of such instruments it is necessary to understand how they instruments affect spatio-temporal land-use patterns. I use game-theoretic and agent-based models to analyse how the generated land-use patterns depend on certain driving forces such as spatially heterogeneous and temporally varying conservation costs and assumptions regarding the behaviour of the agents. These models often form a part of ecological-economic models.

Ecological-economic modelling of conservation policies and strategies

Ecological-economic modelling is an effective instrument for integrating ecological and economic knowledge. I use ecological-economic modelling mainly to develop cost-effective policies and strategies. The ecological-economic models built for this purpose usually consist of an ecological and economic module that assesses the ecological effects and the cost of a policy or strategy. These two are combined within an optimisation framework that takes the institutional setting (principal-agent, multi-agent, uncertainty, information distribution, etc.) into account. I distinguish between planning problems where a planner is perfectly informed about the costs and effects of conservation measures and market-based approaches where information is “asymmetric” such that, e.g., the policy maker has only incomplete information about conservation costs.

Energy modelling

The energy transition in Germany and elsewhere requres the careful expansion of renewable energy production. This has to consider economic costs as well as ecological and societal impact (external effects). A particular research question in this context is the cost-effective and the optimal allocation of renewable power plants as well as the design of potimal market-based instruments for the control of the energy transition.

Optimisation and decision theory

Designing cost-effective land-use strategies and policies generally involves the maximisation of an ecological target function at a given budget constraint (or the minimisation of the budget required for a given ecological target). Such optimisation may be of spatial and/or temporal nature. Furthermore, multiple targets may be present which need to be dealt with in some form of multi-criteria analysis. Often, risk and uncertainty are important phenomena that need to be considered in the optimisation.

Spatial optimisation of conservation strategies
When space plays a role (e.g., in optimal reserve network design) the spatial location of a land-use or conservation measure matters. In a market-based policy the allocation of land-use and conservation measures is controlled by the design of the policy and optimisation means finding the optimal (cost-effective) design. In a planning context I use approaches such as cost-benefit ratio or simulated annealing. The outcome is an ideal solution against which the outcomes of market-based instruments can be compared.

Dynamic optimisation of conservation strategies
Conservation measures can be applied at different points in the course of time. Meadows can, e.g., be mowed at different times during the life cycle of an endangered meadow-dwelling species or at different frequencies. Measures may be made dependent on the ecological (e.g., current population size) or economic (e.g., current budget) situation. Here dynamic programming is the method of choice to solve the optimisation problem. Often, the development of populations and budgets are subject to random influences, which requires the application of stochastic dynamic programming (or approximating heuristics, when the system is too complex for SDP).

Multi-criteria analysis
When land-use or conservation management needs to serve multiple objectives, these may have to be aggregated to obtain a ranking of options. This is particularly demanding when outcomes of measures are subject to uncertainty or when multiple decision makers have to agree on an option.

Dealing with risk and uncertainty
Uncertainty is ubiquitous in environmental systems, which usually reflects in structural and parameter uncertainty in the models used to describe these systems. Sensitivity analysis is sued to assess the impacts of this uncertainty on the outcomes of land-use and conservation strategy. Robustness analysis investigates, e.g., by how much a model (parameter) can be varied to maintain a given rank order of land-use measures. Decision theory provides tools and concepts for such analyses.

Modelling strategies

My background is physical and mathematical modelling, with a preference for mathematical equations where appropriate. Often however, ecological and economic complexity prevents the use of mathematical equations, but rule-based simulation models are the better choice. Recently I developed an interest in agent-based models, as these are very well suited to model human behaviour in a complex environment. Still, I favour agents that are as "simple" as possible and described by only a few essential attributes.

Complex systems

Interactions among agents (particles, plants, animals, humans) may lead to complex dynamics and the formation of distinctive spatio-temporal patterns, such as the clustering of certain attributes in space. Sometimes such patterns can be described by simple laws; sometimes their existence sensitively depends on the interactions among the agents and the environment they are acting within. Complex phenomea make the spice in my research.

Software products

Three software tools have been developed:

  1. PRIMATE - Probabilistic Multi-Attribute Evaluation. I have extended the PROMETHEE method developed by Brans and Mareschal to the consideration of uncertainty and multiple decision makers. The predecessor of PRIMATE has been applied in various stakeholder processes (see Wendy Proctor). You can download the manual. For the free software please send an email.
  2. EcoEcoMod. This software allows to identify cost-effective compensation payments (i.e., payments to compensate farmers for profit losses when they carry out conservation measures) to conserve two endangered butterfly species in a meadow landscape. Users can specify a budget and a mowing regime and watch how the farmers and the butterfly metapopulation respond, 
  3. EcoTRADE. This is a multi-player online game do demonstrate the functioning of tradable permits. Players earn money from using their land for agriculture, but they are obliged to carry out conservation measures on parts of their land. They can meet these targets, or oversupply and sell the excess in the form of certificates to other players, or undersupply and fill the deficit with certificates bought from other players. The goal is to make more money than the other players. 
  4. Soko Bio - SOftware-basierte Entscheidungshilfe zur Bestimmung kosteneffizienter KOmpensationszahlungen für Biodiversitätsschutzmaßnahmen in einer sich ändernden Umwelt. This software helps conservation agencies to design cost-effective compensation payment schemes for the conservation of multiple grassland species.