Production ecology deals with the generation of plant biomass in (agro-) ecosystems as a function of light, water and nutrients. It has important applications in agriculture, where it helps improve production while minimising cost and environmental impact. A thorough understanding of the production ecological principles is needed to describe, explain and optimize production systems. For many crops, dynamic models are developed to simulate crop growth under irrigated and rainfed conditions. There are still major advances made in modelling concepts to better include effects of CO2 fertilization, climate change and weather extremes including droughts. Especially interactions between nutrient supply, drought and plant demand requires further attention to better respond to periods of drought that are expected more frequently for many places. For example, how should nutrient supply be adapted for a season after a season with drought? How will crops respond to a change in nutrient supply, e.g. when applications are split, applied in different forms, at the surface or in the soil near the roots? This requires a detailed understanding of crop nutrient demand and nutrient uptake capacity. For only a few crops, nutrients limitations are included in the dynamic models. However, the concepts still require rigorous testing. In general, crop models are used a lot in science, and they are increasingly used for practical applications in decision-support systems, but improved calibration and validation are needed to allow accurate assessments and decision-support.
Farm system models
At higher levels of integration, farm system models put crop and livestock production into context in terms of management decisions and economic outcomes. These type of models are essential for exploring the interactions between biophysical conditions, production ecological processes and human decisions, actions and requirements. They thereby form powerful integrative tools that allow simulation of what is commonly referred to as Genotype x Environment x Management interaction (GEM). This type of interaction is of great interest to agronomists and plant breeders, since it determines the appropriateness of farming system interventions for different production environments and farm types. Farm system models also allow for integrated assessments, assessing impacts of multiple drivers on multiple impacts, and thereby identifying synergies and trade-offs in pathways towards sustainable development.
Agent-based models (ABMs) are a suitable tool for testing what population- and landscape-level patterns can emerge from the behaviour of interacting individuals. In an ABM, autonomous entities agents (e.g. farmers) and passive objects (e.g. fields) interact with each other and with their environment. ABMs facilitate the simulation of interactions like resource (e.g. biomass, labour) exchanges and cooperation, thereby enabling a population-level analysis that is more than the sum of individual-level functions. Within PPS we use ABMs to explore the effects of diverse scenarios, with various interventions (e.g. sustainable intensification), resource distributions, rules about interaction, and shocks (e.g. climate and market shocks). The model serves to assess potential outcomes of agricultural interventions for individuals and populations, which inform strategic intervention decisions. In various projects we develop ABMs in conjunction with serious games to facilitate discussions with intended beneficiaries about novel ways to interact and remove drivers of unequal outcomes. Ultimately, this leads to the co-design of intervention strategies with the targeted population. A thesis within this sub-theme can contribute to model and game development, but also to testing and refining in collaboration with intended beneficiaries.
Statistical analysis and modelling
Simulation models can describe the expected behaviour of farming systems, but a remaining challenge is posed by the great variability typically observed in production ecological data collected on-farm. This variability reflects both systematic factors related to Genotype x Environment x Management interactions as well as environmental and experimental noise. Separating signal from noise is key to understanding and predicting the on-farm performance of new varieties and technologies in different production environments and farming systems. This requires statistical techniques to analyse the complex and noisy data generated by field trials and surveys performed on-farm. Within this sub-theme you are invited to use and develop innovative methods for the analysis of on-farm, environmental and geospatial data to answer basic questions about the causes and consequences of variation in production ecology, using unique datasets collected from farming systems across the world. Working on this topic could be a great opportunity for anyone interested in production ecology and on-farm agronomy as well as in data science, modelling and statistics.