Ostroski, Anaís
(2024)
Modeling and Optimization Frameworks for Assessing Environmental Impacts and Dependencies of Food Systems at Multiple Spatial Scales.
Doctoral Dissertation, University of Pittsburgh.
(Unpublished)
Abstract
Humans have harnessed animals for thousands of years for a variety of reasons and today animals play crucial roles in food systems. Both their contributions and environmental consequences must be acknowledged for integrated management and moving towards sustainable agricultural practices. On one hand, animal-based products constitute a significant portion of diets. Cattle production has been linked to a series of environmental impacts while occupying almost 41% of land in the contiguous United States between pastureland and cropland for feed production. On the other hand, animal-mediated pollination is critical for production of many food crops. Approximately 90% of flowering plants and at least a third of total crop weight produced globally are dependent on pollinators, mostly wild and managed bees. This work focuses on cattle (animals as products) and bees (animals as resources) as critical avenues to understand interconnections between food systems and the environment. The major goal of the proposed research is two-fold, 1) develop a framework to elucidate the consumption-based environmental impacts (virtual water and nitrogen) of animal-based products via two studies with a specific focus on the US beef supply chain, 2) develop frameworks to assess how agriculture simultaneously impacts and depends on wild bees and their ecosystem services via two studies. For the first goal, modeling and optimization are used to trace and quantify impacts of the beef supply chain. An optimization framework to trace supply chain networks is combined with information on resource use to quantify embodied water and nitrogen emissions associated with beef consumption in the United States. This provides a unique opportunity to understand impacts at the county level, while highlighting clear spatial patterns and discrepancy between production- and consumption-based accounting. Subsequently for the second goal, reported data on soybean yields is coupled with a methodology based on earth observations and machine learning to estimate soybean yield at a fine scale of 250 meters. This fine scale yield data is utilized to map and quantify the interactions between productivity, insecticides, and wild bee abundance across the United States through Bayesian modeling. Finally, a spatial network model of pollination service flows is developed and coupled with an optimization approach to identify areas for bee habitat restoration ultimately pointing towards strategic placement of nesting resources to achieving higher ecosystem service levels in cropland. Overall, this body of work discusses challenges and opportunities for utilizing modeling and optimization for improving the sustainability of food systems.
Share
| Citation/Export: |
|
| Social Networking: |
|
Details
| Item Type: |
University of Pittsburgh ETD
|
| Status: |
Unpublished |
| Creators/Authors: |
|
| ETD Committee: |
|
| Date: |
6 September 2024 |
| Date Type: |
Publication |
| Defense Date: |
2 July 2024 |
| Approval Date: |
6 September 2024 |
| Submission Date: |
18 June 2024 |
| Access Restriction: |
2 year -- Restrict access to University of Pittsburgh for a period of 2 years. |
| Number of Pages: |
190 |
| Institution: |
University of Pittsburgh |
| Schools and Programs: |
Swanson School of Engineering > Civil and Environmental Engineering |
| Degree: |
PhD - Doctor of Philosophy |
| Thesis Type: |
Doctoral Dissertation |
| Refereed: |
Yes |
| Uncontrolled Keywords: |
Environmental impacts, Networks, Water footprint, Nitrogen footprint, Agriculture, Ecosystem services, Pollination |
| Date Deposited: |
06 Sep 2024 19:56 |
| Last Modified: |
06 Sep 2024 19:56 |
| URI: |
http://d-scholarship.pitt.edu/id/eprint/46580 |
Metrics
Monthly Views for the past 3 years
Plum Analytics
Actions (login required)
 |
View Item |
![[img]](https://d-scholarship.pitt.edu/style/images/fileicons/application_pdf.png)
![[feed]](/images/feed-icon-32x32.png){kind=icon}