Lars Caspersen 28/11/2023
As the title already says, we want to investigate nutrient flows for the district of Kleve, Germany. The agricultural sector of Kleve is intense and often export oriented. Animal husbandry (especially dairy cows) plays an important role along side with other intense farming systems like cropping (for example potatos) or ornamental production. Furthermore, Kleve is home to many food processors. A review of substance flow analyses showed, that supply and demand of the agro-food-waste system has become detached (Wiel et al. 2020). This is also the case for Kleve, whose substance flow has become more linear with a large share of imports and exports (Wiel et al. 2021). This repo builds upon the work of Wiel et al. (2021), who mapped the substance flows for nitrogen (N), phosphorous (P), potassium (K) and to a lesser extent carbon (C). We updated the substance flow anaylsis for 2020.
The main goal of this project is to explore socially acceptable solutions of circularity as measure to reduce nitrogen (N) losses and prevent environmental damage. We did that by combining the findings of the (updated) traditional substance flow analysis approach with local perspectives of stakeholders of the agro-food-waste system in Kleve. A schematic representation of the workflow can be seen in the figure blow. In previous studies the large imports of animal feed were identified as a main contributor for reduced nutrient circularity in the district of Kleve (@ Wiel et al. 2021). Based in this observation, we constructed a scenario in which feed imports were inhibited. To understands possible implications of that scenario, we gathered the perspective of local stakeholders via questionnaires and one-to-one interviews. We gathered the input of the stakeholders on four main aspects: 1) allocation of crops to animals, export and biogas substrate 2) composition of livestock herd 3) changes in overall size livestock in Kleve, and 4) allocation of manure to crops, export and manure as biogas substrate. We incorporated the stakeholder perspectives into the scenario. We calculated nutrient flows for the reference year and scenario and based on these calculated circularity indicators.
Here is the answers of the stakeholders on the four aspects of the scenario. Note that most stakeholders were not able to provide their answers in ranges, which is usually the preffered approach for decision analysis studies. It was not possible to conduct stakeholder workshops, so we opted for questionnaires followed up by one-on-one interviews.
The analysis was carried out using a Monte-Carlo simulation with 10000 runs per scenario. We observed a mismatch in the crops allocated to feed the livestock and the expacted changes in livestock size. In the participatory scenario, we did not resolve that mismatch resulting in more nutrients leaving the animal subsystem than entering it. We furthermore explore two ways to resolve the mismatch: 1) ignore the stakeholder proposed allocation of crops to animals and instead allocate as much crops as the estimated herdsize would require (“crop buffered scenario” or “CBS”) or 2) ignore the stakeholder proposed livestock herdsize and reduce the herdsize so that its nutrient demand can be satisfied by the stakeholder-proposed crop allocation (“livestock buffered scenario” or “LBS”). We mapped median changes of most relevant nutrient flows in the figure below, which we adopted from Fernandez et al. (2022).
Based on the modelled nutrient flows, we calculated circularity indicators for the reference year, unaltered participatory scenario, crop buffered scenario, and livestock buffered scenario.
We also mapped the median composition of the total input, the nutrient losses and of the feed allocated to livestock.
The model also provides estimates on potassium (K) and phosphorous (P) nutrient flows, which we decieded to not take into the final analysis, as it was tricky to maintain the stochiometry of the nutrient flows.
Further challenges lied in how to include the stakeholder answers into
the model especially when the answers were clustered in different
intervals. In case the answers are evenly spread, there is no problem in
including the answers in normal distributions as it is usually done in
decision analysis projects using the R package decisionSupport
(Luedeling et al., n.d.). In some cases used normal distributions and in
others we chose skewed normal distributions. For more details on that,
please refer to this github
repository.
Fernandez, Eduardo, Hoa Do, Eike Luedeling, Thi Thu Giang Luu, and Cory Whitney. 2022. “Prioritizing Farm Management Interventions to Improve Climate Change Adaptation and Mitigation Outcomes—a Case Study for Banana Plantations.” Agronomy for Sustainable Development 42 (4): 1–13.
Luedeling, Eike, Lutz Goehring, Katja Schiffers, Cory Whitney, and Eduardo Fernandez. n.d. “Decision Support – Quantitative Support of Decision Making Under Uncertainty.” https://cran.r-project.org/web/packages/decisionSupport/index.html.
Wiel, Bernou Zoë van der, Jan Weijma, Corina Everarda van Middelaar, Matthias Kleinke, Cees Jan Nico Buisman, and Florian Wichern. 2020. “Restoring Nutrient Circularity: A Review of Nutrient Stock and Flow Analyses of Local Agro-Food-Waste Systems.” Resources, Conservation and Recycling 160: 104901.
———. 2021. “Restoring Nutrient Circularity in a Nutrient-Saturated Area in Germany Requires Systemic Change.” Nutrient Cycling in Agroecosystems 121 (2): 209–26.