Progress 03/01/16 to 08/31/21
Outputs
Target Audience:Over the course of the project, we established a network of ranchers in eastern Kansas that have either become informed about the details of the project or partners in the research. Additionally, we connected with land management professionals, and state and federal professionals that assist the ranching community in our region. We also engaged with the Tallgrass Legacy Alliance, which is a rancher-led conservation and management group in our region. Through these various connections, we have connected with numerous specific ranchers that are keenly interested in better ways to manage for this noxious weed in rangelands. Our project has also made important contributions to individuals interested in general principles of environmental sustainability by publishing papers in domain journals in both societies, giving talks in main IE/OR conferences, serving as panel speaker at research and educational panels, and collaborating with researchers from biology, forestry, and environmental sustainability areas. We did not develop a web-based decision tree tool. Initially, we expected that NRCS soil fertility levels and cattle management could be used to tailor herbicide control strategies to the specific landowner situation. As the empirical and modeling approaches matured, it became clear that the Sericea response was too variable across different soil types and cattle management approaches to provide a reasonable forecast of responses to ranch-specific treatment. Indeed, such a tool would be less reliable than implementing principles derived from the general empirical and modeling efforts. Changes/Problems:We were not able to get complete the Sericea detection test because resources had to go to other elements of the project that were deemed more critical. Instead, we used data from a contractor, who does Sericea herbicide work for ranchers in the area, coupled with modeling approaches to examine the potential importance of speed versus thoroughness of search and control strategies of Sericea spread across landscapes. Not only did this approach provide useful insight for Sericea control, but it also provided a useful example of how to handle this problem in other similar weed management contexts. We did not develop a web-based decision tree tool. Initially, expected that NRCS soil fertility levels and cattle management could be used to tailor herbicide control strategies to the specific landowner situation. As the empirical and modeling approaches matured, it became clear that the Sericea response was too variable across different soil types and cattle management approaches to provide reliable guidance. Indeed, such a tool is much less reliable than simply implementing general principles derived from the general empirical and modeling efforts. What opportunities for training and professional development has the project provided?This project has provided a rich combination of training opportunities for students and technicians. For the empirical work, numerous individuals who directly participated in the establishment of the field experiments and data collection efforts. These individuals were a mix of undergraduate, graduate and short-term technicians who received training in ranching practices, weed biology, and general principles of research design. For the modeling component, graduate students have received training in the use of optimization modeling and analysis with applications to invasive species. These experiences include: 1) conducting research in optimization modeling and analysis with applications to invasive species 2) applying the resource allocation research models and solution methods to study practical applications such as controlling Sericea Lespedeza in Kansas 3) presenting research outcomes in two main technical conferences in operations research/industrial engineering and management sciences such as the Institute for Operations Research and the Management Sciences (INFORMS), Production and Operations Management Society (POMS), and Institute of Industrial and Systems Engineers (IISE) research conferences. Outreach Activities: 1) Mentor, Women in OR/MS (WORMS) Coffee and Conversations program, INFORMS Annual Meeting, Phoenix, AZ, 2018 2) Outreach activity and presentation to K-12 Science, Technology, Engineering, Arts, and Math (STEAM) students at Discovery Charter School (scheduled for November 7, 2019) 3) Served as a department representative at the "NCE Graduate Meet and Greet Event," NJIT, October 10, 2019 4) Served as a department representative at the "Academic Showcase, Engineering Open House," NJIT, October 13, 2019 5) Presented at the "Industrial Engineering Academic Information Session," Engineering Open House, NJIT, October 13, 2019 6) Outreach activity and presentation to K-12 Science, Technology, Engineering, Arts, and Math (STEAM) students at Discovery Charter School (scheduled for November 7, 2019) Mentorship of Underrepresented Minorities and Women in STEM: 1) Postdoctoral adviser-Advised, funded, and mentored two female postdoctoral research associates: Ozlem Cosgun and Sevilay Onal (both secured tenure-track Assistant Professor positions at U.S. institutions) 2) Adviser for undergraduate female student Bhumi Patel, who received the 2021 Provost Undergraduate Research and Innovation (URI) Award; advised more than four undergraduate students, resulting in three journal and one conference publications with undergraduate students 3) Mentor, Women in OR/MS (WORMS) Coffee and Conversations program, INFORMS Annual Meeting, Phoenix, AZ, 2018 4) Invited Panel Speaker - IISE Doctoral Colloquium, ISERC 2019, Orlando, FL 5) Invited Panel Speaker - Panel on "Best Practices in Teaching OR/MS", INFORMS 2019, Seattle, WA 6) Outreach activity and presentation to K-12 Science, Technology, Engineering, Arts, and Math (STEAM) students at Discovery Charter School (scheduled for November 7, 2019) 7) Judge, IISE Doctoral Colloquium Best Student Poster Presentation Competition, 2019 8) Judge, INFORMS ENRE student best paper award, 2018 How have the results been disseminated to communities of interest?The PD was invited to give a presentation of the preliminary results to the Tallgrass Legacy Alliance. This group consists primarily of ranchers along with representatives from state and federal agencies and is keenly interested in the results of the project. Additionally, we have communicated the results in numerous contexts including: The CO-PD has organized and chaired the Energy, Natural Resources and the Environment (ENRE) sponsored session, titled "Spatio-Temporal Conservation Models for Controlling Biological Invasions" in the 2017 INFORMS Annual Meeting in Houston, TX and presented results and findings of this research at this ENRE session. The CO-PD has been invited to present her research on "A Mixed-Integer Programming Approach to the Parallel Replacement Problem under Technological Change" at the INFORMS MIF Best Paper Competition Session at the 2017 INFORMS Annual Meeting (The PI's research publication has been selected as a finalist at the 2017 INFORMS MIF Best Paper Competition) Other presentations not listed under "products" I. Esra Büyüktahtakin, "Controlling Biological Invasions and Epidemics," NJIT Faculty Showcase, March, 2018. Najmaddin Akhundov*, I. Esra Büyüktahtakin, Jennifer Smith, Gregory R. Houseman, "An integrated simulationoptimization framework to optimizing search and treatment path for controlling a biological invader," Dana Knox Student Research Showcase, NJIT, 2018 (Najmaddin Akhundov, former Ph.D. student in SOAL, received first place among 2018 graduate researchers at Dana Knox Student Research Showcase, NJIT, 2018). The Co-PD gave a seminar on "Problem Solving & Decision Making" to middle school students at a STEM camp in Wichita. She also had a brief presentation about "Industrial and Systems Engineering and Applications to Invasive Species Management" at the Wallace Invitational for Scholarships in Engineering (WISE) Panel organized by the Society of Women Engineers (SWE) and College of Engineering at Wichita State University. Invited Seminars or Panels: 1. I. E. Büyüktahtakin (Invited Panel Speaker), "IISE Doctoral Colloquium," ISERC 2019, Orlando, FL. 2. I. E. Büyüktahtakin (Invited Panel Speaker), "NJIT Panel Discussion on NSF CAREER Award," NJIT, October 2018. 3. I. E. Büyüktahtakin has been invited to give a seminar on "Stochastic Mixed-Integer Programming Approaches to the Optimal Surveillance and Control of Biological Invasions," in Industrial and Manufacturing Systems Engineering, Kansas State University, March, 2017. What do you plan to do during the next reporting period to accomplish the goals?
Nothing Reported
Impacts
What was accomplished under these goals?
For the empirical work (Goal A), we undertook a massive effort to collect an unprecedented level of population level data on Sericea under highly realistic conditions over multiple years. Specifically, we established 53, 1-acre monitoring plots across 7 active cattle ranches in the Flint Hills of Kansas. The plots represented differences in cattle management, soil fertility, and frequency of herbicide application (1, 2, or 3-year intervals). In total, we marked and followed the fates over 2000 Sericea individuals. We also collected seeds for estimating reproductive capacity and quantified new germinants over several years. After four growing seasons, there was no evidence for a consistent effect of soil fertility on Sericea density (X2=0.14, p=0.9). Annual herbicide application reduced ramet density to approximately a third of the three-year application internal (X2=11.7, p<0.003). Cattle management had inconsistent effects on Sericea density alone or in combination with other factors. Seed production varied widely among plots and was strongly impacted by herbicide treatment interval which reduced any strong relationships with soil fertility or cattle management. Seed type was relatively consistent across years and treatments ranging from 63-87% cleistogamous (selfed) seed. Although the seed number increased with plant size (number of ramets), this relationship was very weak and variable suggesting that Sericea shifts to rhizomatous spread rather than increasing seed production on individual ramets under cattle grazing. Seedling emergence was influenced by cattle management (X2=20.1, p<0.0002), and year of observation (X2=30.6, p<0.0001) and soil fertility depending on the specific year (X2=12.6, p=0.013). Seedling emergence decreased with each year which would be expected for pastures with consistent herbicide control measures. Taken together, soil fertility appears to have weak effects on the spread of Sericea relative to cattle management and intervals of herbicide application. For the cattle management strategies examined, some may have greater vulnerability to Sericea spread (harder to contain), but these effects were quite variable. One of the aspirations of the empirical work was to find a new weakness or vulnerability in the biology of Sericea that we might use to reduce or improve efficiency of herbicides. Unfortunately, no clear vulnerability emerged. However, we did find evidence from lab germination trials that the seeds from chasmogamous (outcrossed) flowers have low dormancy while the cleistogamous (selfed) flowers have much higher dormancy and likely contribute strongly to the long-lived seedbank which is a pernicious problem that prevents Sericea eradication once a population is established. This novel result is relevant to improved ecological understanding of plants with a mixed-mating system. The research under Goal B Objectives 2, 3, and 4 resulted in the paper (Onal et al. 2020). In this paper, the research team has developed a novel integrated simulation-optimization framework to effectively search and treat invasive species under a limited management budget. The simulation mimics invader growth over a landscape and 12 years over a 100-acre landscape, while a bio-economic optimization model finds an optimal search and treatment path to minimize its economic damage to agricultural production. None of the former models in the literature of invasive species management has considered the search and treatment path optimization problem before, and thus our optimization model is novel. The advantage of our simulation-optimization approach compared to the single bio-economic optimization model of Büyüktahtakin et al. (2015) is that we are able to solve the simulation and optimization models consecutively, resulting in a significant reduction in problem complexity and solution time. Furthermore, our optimization model determines the optimal path for searching and treating invasive species, which has not been considered formerly in the literature. Our first experimental objective determined if the annual, biennial or triennial treatment provided the most effective treatment strategy. Based on our computations, applying the treatment every year results in the lowest economic damage under each scenario. We also found that the initial frequency had a higher effect on the treatment costs than the abundance patterns in the sites occupied. However, the highest benefit from treatment is obtained under high abundance rather than high frequency. The second key finding was that slow speed treatment strategy with the highest effectiveness rate is the best treatment option, as observed in seven out of the nine cases in our experimental study. On the other hand, experiments on the low initial abundance case suggest using an intermediate speed for search and treatment unless the initial frequency is high. Thus, our results imply that applying yearly treatment with a slow treatment speed results in the biggest bang for the buck under most invasion scenarios. The research under Goal B Objectives 2, 3, 4, and 5 has also resulted in the paper (Onal et al. in prep. 2022) and several conference presentations. In this paper, we have presented an integrated simulation-optimization model that estimates the seed dispersal using Gaussian cell-to-cell transition probabilities. Different than the former model (Onal et al. 2020), here our new model prescribes the treatment over a 25-year time period and a 2500-acre landscape. To our knowledge, this is the largest spatial and temporal simulation of invasive species control that has been performed in the literature. In this study, our main accomplishment is to present a unified, automated simulation-optimization problem that can run for any landscape size, time period, and combination of parameters. Different than former models, this simulation framework considers unexpected dispersal. Thus, a very important advancement is the incorporation of unexpected spread of an invader by using Gaussian seed dispersal over a large landscape. This model assumes a center-to-center distribution for each cell in an area, calculates the probability of a seed traveling from one cell to every other cell by calculating the Euclidean distances between cell centers, and normalizing them by dividing them by the sum of the probability densities of two placement kernels. Furthermore, we have included an additional uniform distribution of seed dispersal to capture random dispersal, that could not be explained with any known outside factors, such as animal or vehicle movement. Our main findings indicate that the optimal treatment strategies are quite sensitive to the input data, such as the landscape size and the planning horizon, and depend on several factors, such as the budget, the germination rate, and dispersal rates. Thus, it is difficult to find intuitive strategies without using a complex mathematical model such as ours, implying the need to utilize a simulation-optimization mathematical modeling framework to capture the most effective management strategies to combat the spread of Sericea. Summary: Our empirical patterns and modeling efforts clearly demonstrate that consistent herbicide management is needed to contain the spread of Sericea, while poor or inconsistent herbicide control will lead to exponential increases in the problem. Slow and through search and herbicide application is more likely to slow Sericea spread than trying to work quickly to cover more area each year. Additionally, the substantial advancement in optimization modeling accomplished holds promise for addressing complex land management decision far beyond Sericea control in rangelands.
Publications
- Type:
Theses/Dissertations
Status:
Published
Year Published:
2018
Citation:
Smith, J. N. 2018. Population demographics of an invasive weed: responses to soil fertility and cattle management. Thesis, Wichita State University, Wichita, Kansas, U.S.A.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2017
Citation:
Smith, Jennifer N. 2017. Assessing population demographics of sericea (Lespedeza cuneata) in response to environmental variation and land management in the Great Plains--In Proceedings: 13th Annual Symposium on Graduate Research and Scholarly Projects. Wichita, KS: Wichita State University.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2018
Citation:
Sevilay Onal*, Najmaddin Akhundov*, I. Esra B�y�ktahtakin, Jennifer Smith, Gregory R. Houseman, An integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader, INFORMS Annual Meeting, November 2018, Phoenix, AZ.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
Najmaddin Akhundov*, Sevilay Onal*, Esra B�y�ktahtakin, Jennifer Smith, Gregory R. Houseman, An integrated simulation-optimization framework to optimize search and treatment path for controlling a biological invader, INFORMS Computing Society Conference, January 2019, Knoxville, TN.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
Sevilay Onal*, Najmaddin Akhundov*, I. Esra B�y�ktahtakin, Jennifer Smith*, Gregory R. Houseman, Detection and Control Operations Management of an Agricultural Invader Through Integrated Simulation-Optimization, POMS Conference, May 2019, Washington, DC.
- Type:
Journal Articles
Status:
Published
Year Published:
2020
Citation:
Sevilay Onal*, Najmaddin Akhundov*, I. Esra B�y�ktahtakin, Jennifer Smith*, Gregory R. Houseman. 2020. An integrated simulation-optimization framework to optimize search and treatment path for controlling a biological invader. International Journal of Production Economics, Volume 222, April 2020, 107507, https: //doi.org/10.1016/j.ijpe.2019.09.028
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2019
Citation:
Sevilay Onal*, I. Esra B�y�ktahtakin, Sabah Bushaj, Jennifer Smith*, Gregory R. Houseman Gaussian Dispersal Model for Controlling a Biological Invader, INFORMS Conference, Oct 2019, Seattle, WA.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2020
Citation:
Sevilay Onal*, Sabah Bushaj*, Esra B�y�ktahtak, Jennifer Smith*, Gregory R. Houseman Gaussian Dispersal Model For Controlling A Biological Invader, INFORMS Virtual Conference, Oct 2020.
|
Progress 03/01/20 to 02/28/21
Outputs
Target Audience:The PI continues to work with the Tallgrass Legacy Alliance, an organization focused on ranching issues in the Flint Hills of Kansas.To this point, dissemination of results is limited because the data do not yet support any new recommendations.However, the research has been discussed with numerous students involved with the project and in courses taught by the PI.Likewise, the PI has continued to work with specific ranchers that are keenly interested in better ways to manage for this noxious weed in rangelands. During the fifth year of the project, the CO-PD has continued her efforts on cultivating synergy among academics in OR and Environmental Sustainability by publishing papers in domain journals in both societies, giving talks in main IE/OR conferences (to be given in April of 2021). Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?The establishment of the empirical work relied heavily on assistance from undergraduate and technicians who gained experience in fieldwork, data collection, and scientific approaches to complex problems. Under the supervision of the PDs, students received training in ranching practices, weed biology, and general principles of research design. For the modeling component, graduate students have received training in the use of optimization modeling and analysis with applications to invasive species. These experiences include: 1) conducting research in optimization modeling and analysis with applications to invasive species 2) applying the resource allocation research models and solution methods to study practical applications such as controlling Sericea Lespedeza in Kansas 3) presenting research outcomes in two main technical conferences in operations research/industrial engineering and management sciences such as the Institute for Operations Research and the Management Sciences (INFORMS), Production and Operations Management Society (POMS), and Institute of Industrial and Systems Engineers (IISE) research conferences. How have the results been disseminated to communities of interest?The primary dissemination has been through individual interactions with landowners, researchers, and professionals. The results are preliminary at this point so great care must be taken to not misrepresent the certainty of data. What do you plan to do during the next reporting period to accomplish the goals?We have been given an extension for the project herbicide experimental plots for landowners who preferred that we perform a final round of chemical control following the experiment. In the several months, the PI's plan to: 1) Continue developing the search and control model based on the Gaussian dispersion model on a larger landscape level and analyze tradeoffs between search & treatment efforts under Goal B Objectives 3 & 4 using the 4-years of data provided by the PD Houseman. 2) Continue manuscript preparation for a peer-reviewed journal. 3) Continue the efforts in disseminating research findings through venues of leading conferences, providing seminars to collaborators in natural resources conservation and environmental sustainability as well as communities of interests, and providing opportunities for training and professional development for students.
Impacts
What was accomplished under these goals?
Goal A The primary empirical effort in the fifth of the project was completion of data collection from 26 plots across 6 ranches within the Flint Hills region of Kansas. We rechecked marked plants from previous censuses to quantify survivorship. Following this last year of data collection, we removed all of our marking devices (t-posts, landscape nails, etc). We are compiling the field data for incorporation into different analytical products. Goal B The research under Goal B Objectives 2, 3, and 4 resulted in the paper (Sevilay Onal*, Najmaddin Akhundov*, I. Esra Büyüktahtakin et al. (2020); that has been published in International Journal of Production Economics). In this paper, the research team have developed a new integrated simulation-optimization model to effectively search and treat invasive species under a limited management budget. The simulation mimics invader growth over a landscape and 12 years, while a bio-economic optimization model finds an optimal search and treatment path to minimize its economic damage to agricultural production. Our optimization model is new in the sense that it prescribes the optimal path for searching and treating sites for controlling the invader. This study is also the first in the literature to present and combine simulation and optimization models of the complex bio-economic search-path problem, and to prescribe a pathway for search and treatment that is applicable to the real-world management of invasive species. Our case study data and parameter calibration are based on the large-scale field data on Sericea, an aggressive invasive plant threatening the Great Plains of the U.S., collected over twelve pastures in Kansas during the past two years. Solving simulation and optimization models in a consecutive fashion provides a considerable computational advantage to find an optimal solution to practical size problems in a reasonable time. Our results imply that applying yearly treatment with a slow search-and-treatment speed results in the biggest bang for the buck under most invasion scenarios. Under research under Goal B Objectives 3&4, the research team has been working on developing an integrated simulation-optimization model, which will incorporate long-distance dispersal to optimize resource allocation between search and treatment. Using this integrated simulation-optimization model, we will study tradeoffs between detecting new random emergence of individuals versus treating older already-detected patches. Different from the former paper (Onal et al., 2020), the computational framework will automize the simulated growth of the invader and the optimization of resource allocation. During the fifth year, the PhD student and the post-doc spent considerable amount of time to write a code to automize the runs of simulation and optimization models. A preliminary model of the long-distance dispersal has been formulated based on the Gaussian dispersion model. The simulation model attempts to include more realistic long-distance dispersal and the Gaussian model, while the optimization model is simplified to determine only the locations for search and control, rather than a search path, in order to apply the model to larger landscapes and longer time horizons compared to the model presented in Onal et al. (2020). The model will be calibrated using 4-years of data collected over seventeen pastures in Kansas.
Publications
|
Progress 03/01/19 to 02/29/20
Outputs
Target Audience:The PI continues to work with the Tallgrass Legacy Alliance, which is a ranching organization focused on issues in the Flint Hills of Kansas.To this point dissemination of results is limited because the data do not yet support any new recommendations.However, the research has been discussed with numerous students involved with the project, in courses taught by the PI, and with guests taken to one of the field sites.Likewise, the PI has worked with specific ranchers that are keenly interested in better ways to manage for this noxious weed in rangelands. The CO-PD Buyuktahtakin has worked with and funded a female post-doctoral student and a doctoral student in research efforts. The CO-PD provided students with opportunities to present their research at national, state, and university level conferences. The post-doctoral student Sevilay Onal, has joined University of Illinois at Springfield as an Assistant Professor of Business Administration in August 2019. During the fourth year of the project, the CO-PD has continued her efforts on cultivating synergy among academics in Operations Research and Environmental Sustainability by publishing papers in domain journals in both societies, giving talks in main IE/OR conferences, serving as panel speaker at research and educational panels, and collaborating with researchers from biology, forestry, and environmental sustainability areas. The details of those dissemination efforts and talks are given below under the "Products" section. Changes/Problems:As described last year, the cost of working across multiple ranches has led to much higher expenditures for travel and personnel than anticipated. This investment has been highly valuable, but has limited what we can do for some aspects of the project. Specifically, we will not be able to quantify the Sericea detection and treatment probabilities (objective 3). Objectives 1-2 have been addressed and each is far more important to the success of the overall project. To overcome the lack of data from objective 3, we are working with a commercial applicator to ascertain time allocation for his search and control efforts on one of the ranches. Using this information, we plan to include this dimension in the modeling efforts. At the outset of the project, the cooperating landowners agreed to refrain from aerial spraying of the experimental areas during the project. Unfortunately, some landowners changed their minds and have started aerial spraying. This means that we will not have data for all years for some of our plots. What opportunities for training and professional development has the project provided?The establishment of the empirical work relied heavily on assistance from graduate and undergraduate students. Under the supervision of the PDs, students received training in ranching practices, weed biology, and general principles of research design. For the modeling component, graduate students have received training in the use of optimization modeling and analysis with applications to invasive species. These experiences include: 1) conducting research in optimization modeling and analysis with applications to invasive species 2) applying the resource allocation research models and solution methods to study practical applications such as controlling Sericea Lespedeza in Kansas 3) presenting research outcomes in two main technical conferences in operations research/industrial engineering and management sciences such as the Institute for Operations Research and the Management Sciences (INFORMS), Production and Operations Management Society (POMS), and Institute of Industrial and Systems Engineers (IISE) research conferences. How have the results been disseminated to communities of interest?The primary dissemination has been through individual interactions with landowners, researchers, and professionals. The results are highly preliminary at this point so great care must be taken to not misrepresent the certainty of data. What do you plan to do during the next reporting period to accomplish the goals?In the next year, PD Houseman and his team plans to complete data collection to refine the bioeconomic model and quantify the demography of the species. In the next year of research, the CO-PD plans to: 1. Continue to developing the search and control model based on the Gaussian dispersion model on a larger landscape level and analyze tradeoffs between search and treatment efforts under Goal B Objectives 3 & 4 using the 4-years of data provided by the PD Houseman. 2. Continue to work on the paper under preparation for journal submission. 3. Continue the efforts in disseminating research findings through venues of leading conferences, providing seminars to collaborators in natural resources conservation and environmental sustainability as well as communities of interests, and providing opportunities for training and professional development for students.
Impacts
What was accomplished under these goals?
Goal A The primary empirical effort in the fourth year of the project was data collection from 26 plots across 6 ranches within the Flint Hills region of Kansas. We rechecked individual marked plants from previous censuses and, if dead, a replacement was marked for further quantification of survivorship. We also collected seed germination data and community biomass production at each plot. Unfortunately, we have lost a number of plots because some landowners aerial sprayed their properties despite previous assurances that no aerial spraying would be conducted during the duration of our experiment. Nevertheless, we expect to have a robust dataset once we have the data for all years and all available plots. Goal B Under Goal B Objective 3, the CO-PD Buyuktahtakin and her team have developed an integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader. The fourth year research activities under Under Goal B Objective 3 have resulted in the following paper under review for International Journal of Production Economics, 2019 (* indicates student author): Sevilay Onal*, Najmaddin Akhundov*, I. Esra Büyüktahtakin, Jennifer Smith*, Gregory R. Houseman, "An integrated simulation-optimization framework to optimize search and treatment path for controlling a biological invader," International Journal of Production Economics, Volume 222, April 2020, 107507, https://doi.org/10.1016/j.ijpe.2019.09.028 The aim of Goal B Objective 4 is to evaluate the effectiveness of the optimization model strategy to combat ongoing or emerging weed management problems. The CO-PD has tested and evaluated predictions of the search and control model using data from the literature and simulation models for sericea invasion in the Great Plains. Collaborator Dr. Greg Houseman from Wichita State Biological Sciences has collected significant data from his field work. This data includes sericea germination rate, sericea seed production, stem density, sericea stem frequency over the landscape, and sericea survival rate from multiple pastures in Kansas. We have analyzed 2 years of field-based data and used it for parameter calibration and model validation for the search and control model. We will refine the input data after obtaining the third and fourth year observational data from the field. The research under Goal B Objectives 2, 3, and 4 resulted in the paper (Sevilay Onal*, Najmaddin Akhundov*, I. Esra Büyüktahtakin et al. (2020); that has been published in International Journal of Production Economics). In this paper, the research team have developed a new integrated simulation-optimization model to effectively search and treat invasive species under a limited management budget. The simulation mimics invader growth on a landscape over 12 years, while a bio-economic optimization model finds an optimal search and treatment path to minimize its economic damage to agricultural production. Our optimization model is new in the sense that it prescribes the optimal path for searching and treating sites for controlling the invader. This study is also the first in the literature to combine simulation and optimization models of the complex bio-economic search-path problem, and to prescribe a pathway for search and treatment that is applicable to the real-world management of invasive species. Our case study data and parameter calibration are based on the large-scale field data on Sericea, an aggressive invasive plant threatening the Great Plains of the U.S. Solving simulation and optimization models in a consecutive fashion provides a considerable computational advantage to find an optimal solution to practical size problems in a reasonable time. Our results imply that applying yearly treatment with a slow search-and-treatment speed results in the biggest bang for the buck under most invasion scenarios. Under research under Goal B Objectives 3&4, the research team has been working on developing an integrated simulation-optimization model, which will incorporate long-distance dispersal to optimize resource allocation between search and treatment. Using this integrated simulation-optimization model, we will study tradeoffs between detecting new random "pop-up" patches and treating previously detected patches. Different from the former paper (Onal et al., 2020), the computational framework will automize the simulated growth of the invader and the optimization of resource allocation. During the fourth year, the PhD student and the post-doc spent considerable time writing code to automize the runs of simulation and optimization models. A preliminary model of the long-distance dispersal has been formulated based on the Gaussian dispersion model. The simulation model accounts for long distance dispersal and the Gaussian model, while the optimization model is simplified to determine only the locations for search and control, rather than a search path, in order to apply the model to larger landscapes and longer time horizons compared to the model presented in Onal et al. (2020). The model will be calibrated using 4-years of data collected.
Publications
- Type:
Journal Articles
Status:
Published
Year Published:
2020
Citation:
Sevilay Onal*, Najmaddin Akhundov*, I. Esra B�y�ktahtakin, Jennifer Smith*, Gregory R. Houseman, An integrated simulation-optimization framework to optimize search and treatment path for controlling a biological invader, International Journal of Production Economics, Volume 222, April 2020, 107507, https://doi.org/10.1016/j.ijpe.2019.09.028
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2020
Citation:
Sevilay Onal*, I. Esra B�y�ktahtakin, Sabah Bushaj, Jennifer Smith*, Gregory R. Houseman Gaussian Dispersal Model for Controlling a Biological Invader, INFORMS Conference, Oct 2019, Seattle, WA.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2020
Citation:
Sevilay Onal*, Najmaddin Akhundov*, I. Esra B�y�ktahtakin, Jennifer Smith*, Gregory R. Houseman, Detection and Control Operations Management of an Agricultural Invader Through Integrated Simulation-Optimization, POMS Conference, May 2019, Washington, DC.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2020
Citation:
Sevilay Onal*, Najmaddin Akhundov*, I. Esra B�y�ktahtakin, Jennifer Smith, Gregory R. Houseman, An integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader, INFORMS Annual Meeting, November 2018, Phoenix, AZ.
|
Progress 03/01/18 to 02/28/19
Outputs
Target Audience:The PI continues to work with the Tallgrass Legacy Alliance, which is a ranching organization focused on ranching issues in the Flint Hills of Kansas.To this point, dissemination of results is limited because the data do not yet support any new recommendations. The CO-PD Buyuktahtakin has identified and included one female post-doctoral student in research efforts. The CO-PD has funded one doctoral student and one female postdoctoral research associate under this research. The CO-PD provided students with opportunities to present their research at national, state, and university level conferences. The post-doctoral student Sevilay Onal, has received a tenure-track faculty position offer from the University of Illinois at Springfield, and will be joining the University of Illinois at Springfield as an Assistant Professor of Business Administration in August 2019. During the third year of the project, the CO-PD has continued her efforts on cultivating synergy among academics in OR and Environmental Sustainability by publishing papers in domain journals in both societies, giving talks in main IE/OR conferences, serving as panel speaker at research and educational panels, and collaborating with researchers from biology, forestry, and environmental sustainability areas. The details of those dissemination efforts and talks are given below under the "Products" section. Changes/Problems:As stated last year, the travel costs in terms of mileage and people time has exceeded projections. Our initial proposal did not include multiple ranches, but in response to constructive feedback we have done our best to get multiple sites included in the design. This was a wise move, but the additional costs will force us to treat year 4 as the final year of comprehensive data collection. Although more years are always advantageous, we think that the data collected over four field seasons will be sufficient to address the core goals of the project. What opportunities for training and professional development has the project provided?The establishment of the empirical work relied heavily on assistance from graduate and undergraduate students. Under the supervision of the PDs, students received training in ranching practices, weed biology, and general principles of research design. For the modeling component, graduate students have received training in the use of optimization modeling and analysis with applications to invasive species. These experiences include: 1) conducting research in the areas of optimization modeling and analysis with applications to invasive species 2) applying the resource allocation research models and solution methods to study practical applications such as controlling Sericea Lespedeza in Kansas 3) presenting research outcomes in two main technical conferences in operations research/industrial engineering and management sciences such as the Institute for Operations Research and the Management Sciences (INFORMS), Production and Operations Management Society (POMS), and Institute of Industrial and Systems Engineers (IISE) research conferences. How have the results been disseminated to communities of interest?The primary dissemination has been through individual interactions with landowners, researchers, and professionals. The results are highly preliminary at this point so great care must be taken to not misrepresent the certainty of data. The CO-PD Buyuktahtakin and her students have been disseminating research findings through journal paper publications, organizing and chairing sessions at the Institute for Operations Research and the Management Sciences (INFORMS) conference, and collaborating with professionals from practice. The CO-PD and her team who worked on this project have presented technical papers for invasive species control and optimization research at the 2018 INFORMS Annual Meeting, at the 2019 INFORMS Computing Society Conference that was held in Knoxville, TN in January 2019, and the 2019 POMS Conference, which was held in May 2019, in Washington, DC. In the third year of research, the CO-PD has actively disseminated the research results through three major conference venues by presenting technical papers and organizing special sessions related to optimization and invasive species management. In detail: Conference Presentations: 1. Sevilay Onal*, Najmaddin Akhundov*, I. Esra Büyüktahtakin, Jennifer Smith*, Gregory R. Houseman, "Detection and Control Operations Management of an Agricultural Invader Through Integrated Simulation-Optimization," POMS Conference, May 2019, Washington, DC. 2. Najmaddin Akhundov*, Sevilay Onal, I. Esra Büyüktahtakin, Jennifer Smith, Gregory R. Houseman, "An integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader," INFORMS Computing Society Conference, January 2019, Knoxville, TN. 3. Najmaddin Akhundov*, I. Esra Büyüktahtakin, Jennifer Smith, Gregory R. Houseman, "An integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader," Dana Knox Student Research Showcase, NJIT, 2018 (Najmaddin Akhundov, former Ph.D. student in SOAL, received first place among 2018 graduate researchers at Dana Knox Student Research Showcase, NJIT, 2018). 4. I. Esra Büyüktahtakin, "Controlling Biological Invasions and Epidemics," NJIT Faculty Showcase, March, 2018. Invited Seminars or Panels: 1. I. E. Büyüktahtakin (Invited Panel Speaker), "IISE Doctoral Colloquium," ISERC 2019, Orlando, FL. 2. I. E. Büyüktahtakin (Invited Panel Speaker), "NJIT Panel Discussion on NSF CAREER Award," NJIT, October 2018. Poster Presentations: 1. Sevilay Onal*, Najmaddin Akhundov*, I. Esra Büyüktahtakin, Jennifer Smith, Gregory R. Houseman, "An integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader," INFORMS Annual Meeting, November 2018, Phoenix, AZ. What do you plan to do during the next reporting period to accomplish the goals?In the next year, PD Houseman and his team plans to continue data collection necessary to refine the bioeconomic model and quantify the demography of the species. In the next year of research, the CO-PD plans to: 1. continue to developing the search and control model on a larger landscape level and analyze tradeoffs between search & treatment efforts under Goal B Objectives 3 & 4 using the 4-year data provided by the PD Houseman. 2. continue to work on the papers that were submitted to journals and under preparation. 3. continue the efforts in disseminating research findings through venues of leading conferences, providing seminars to collaborators in natural resources conservation and environmental sustainability as well as communities of interests, and providing opportunities for training and professional development for students.
Impacts
What was accomplished under these goals?
Goal A The primary empirical effort in the third year of the experiment was data collection from 38 plots across 7 ranches within the Flint Hills region of Kansas. We rechecked each surviving individual from 2017-2018 census and, if dead, a replacement was marked for further quantification of survivorship. Additionally, we are collecting seed germination data and community biomass production. These data are pivotal for understanding spatio- and temporal variation among Sericea populations. As for survivorship, it appears that each life stage was similarly affected by herbicide application in 2018. This was not the case in 2017 when earlier life stages had greater survivorship than larger/older individuals. Once again, this illustrates the importance of multi-year investigations. Currently we are attempting to update the projected population growth rates based on the additional data. Goal B Under Goal B Objective 3, the CO-PD Buyuktahtakin and her team have developed an integrated simulation-optimization framework to optimizing search and treatment path for controlling a biological invader. The third year research activities under Under Goal B Objective 3 have resulted in the following paper under review for International Journal of Production Economics, 2019 (* indicates student author): Sevilay Onal*, Najmaddin Akhundov*, I. Esra Büyüktahtakin, Jennifer Smith*, Gregory R. Houseman, "An integrated simulation-optimization framework to optimize search and treatment path for controlling a biological invader," under review for International Journal of Production Economics, 2019. Invasive species cause large economic losses and are difficult to control due to the high cost of locating and treating the continual emergence of new individuals. The national strategy against invasive species encompasses prevention, early detection, and rapid response. In particular, determining an optimal route for search and treatment under limited budget is critical to reducing management costs. In this paper (Onal et al., 2019; under review), we present a new integrated simulation-optimization framework (see Figure 1 below) to effectively search and treat invasive species under a limited management budget. The simulation mimics invader growth over a landscape and 12 years, while a bio-economic optimization model finds an optimal search and treatment path to minimize its economic damage to agricultural production. This study is the first in the literature to present and combine simulation and optimization models of the complex bio-economic search-path problem, and to prescribe a pathway for search and treatment that is applicable to the real world management of invasive species. Our case study data and parameter calibration are based on the large-scale field data on Sericea, an aggressive invasive plant threatening the Great Plains of the U.S., collected over eighteen pastures in Kansas during the past two years. Solving simulation and optimization models in a consecutive fashion provides a considerable computational advantage to find an optimal solution to practical size problems in a reasonable time. Our results imply that applying yearly treatment with a slow search-and-treatment speed results in the biggest bang for the buck under most invasion scenarios. The aim of Goal B Objective 4 is to evaluate the effectiveness of the optimization model strategy to combat ongoing or emerging weed management problems. The CO-PD has tested and evaluated predictions of the search and control model using data from the literature and simulation models for sericea invasion in the Great Plains. Collaborator Dr. Greg Houseman from Wichita State Biological Sciences has collected significant data from his field work. This data includes sericea germination rate, sericea seed production, stem density, sericea stem frequency over the landscape, and sericea survival rate from multiple pastures in Kansas (e.g., Brink pasture, Vestring Pasture, Youngmeyer Pasture, Stotts pasture, Browning pasture). We have analyzed 2 years of field-based data, and used it for parameter calibration and model validation for the search and control model. We will refine the input data after obtaining the third and fourth year observational data from the field.
Publications
|
Progress 03/01/17 to 02/28/18
Outputs
Target Audience:During Fall 2017, the CO-PD has offered a new graduate course titled "IE 705: Mathematical Programming in Management Science". In this course the CO-PD has integrated the research and educational efforts by including a significant term project that involves the implementation of an optimization problem in environmental management and sustainability. Students taking this class have had the opportunity to work on problems related to invasive species management, sericea search and control problem, and carbon sequestration network optimization. During the second year of the project, the CO-PD has continued her efforts on cultivating synergy among academics in operations research and environmental sciences by publishing papers in domain journals in both societies, organizing sessions and giving talks in main industrial engineering/operations research conferences, serving as panelists at CAREER panels, and collaborating with researchers from biology, forestry, and environmental sustainability areas. Changes/Problems:One unexpected development in this second year was that a few areas containing 12 plots were aerial sprayed with herbicide. This occurred despite assurances from landowners at the outset that they would not spray these areas during our project. We are currently considering how to adapt our design to this unexpected turn of events. A second problem is that the GPS-sprayer apparatus that we planned to use for the detection experiment has not worked as planned. We are currently re-examining how to best address this problem. Finally, the travel costs in terms of mileage and people time has exceeded projections. Our initial proposal did not include multiple ranches, but in response to constructive feedback from the review panel we have done our best to get multiple sites included in the design. This was a wise move, but that change made it very difficult to estimate the costs. It may become necessary in years 4-5 to adjust our sampling to meet budget limitations. What opportunities for training and professional development has the project provided?The establishment of the empirical work relied heavily on assistance from a graduate studentand field technicians. Under the supervision of the PDs, students received training in ranching practices, weed biology, and general principles of research design. For the modeling component, graduate students have received training in the use of optimization modeling and analysis with applications to invasive species. These experiences include: 1) conducting research in the areas of optimization modeling and analysis with applications to invasive species 2) applying the resource allocation research models and solution methods to study practical applications such as controlling Sericea Lespedeza in Kansas 3) presenting research outcomes in two main technical conferences in operations research/industrial engineering and management sciences such as the Institute for Operations Research and the Management Sciences (INFORMS) and Institute of Industrial and Systems Engineers (IISE) research conferences. How have the results been disseminated to communities of interest?The primary dissemination has been through individual interactions with landowners, researchers, and professionals. The results are highly preliminary at this point so great care must be taken to not misrepresent the certainty of the results. The CO-PD and her students have been disseminating research findings through conference presentations, organizing and chairing sessions at the Institute for Operations Research and the Management Sciences (INFORMS) conference, and collaborating with professionals from practice. Specifically: The CO-PD has organized and chaired the Energy, Natural Resources and the Environment (ENRE) sponsored session, titled "Spatio-Temporal Conservation Models for Controlling Biological Invasions" in the 2017 INFORMS Annual Meeting in Houston, TX and presented results and findings of this research at this ENRE session. The CO-PD has been invited to present her research on "A Mixed-Integer Programming Approach to the Parallel Replacement Problem under Technological Change" at the INFORMS MIF Best Paper Competition Session at the 2017 INFORMS Annual Meeting (The PI's research publication has been selected as a finalist at the 2017 INFORMS MIF Best Paper Competition) I. E. Büyüktahtakin has been invited to give a seminar on "Stochastic Mixed-Integer Programming Approaches to the Optimal Surveillance and Control of Biological Invasions," in Industrial and Manufacturing Systems Engineering, Kansas State University, March, 2017. Due to the PI's research accomplishments, the PI was given the WSU Young Faculty Scholar Award by Wichita State University in May 2017. The PI's research paper titled "A Mixed-Integer Programming Approach to the Parallel Replacement Problem under Technological Change" has also been selected as a finalist at the 2017 INFORMS MIF Best Paper Competition. The PI has also been invited to present her research at INFORMS MIF Best Paper Competition Session at the 2017 INFORMS Annual Conference held in Houston, TX. What do you plan to do during the next reporting period to accomplish the goals?In the next year, PD Houseman and his team plans to continue data collection necessary to refine the bioeconomic model and quantify the demography of the species. He also plans to refine the approach for quantifying the probability of Sericea detection. Additionally, the team will continue to developing the search and control model in Goal B Objective 3 using the 3-year data provided by the PD Houseman. In particular, the CO-PD will revise the search path model into a spatial search and control model in which tradeoffs between allocating resources among search efforts to detect new populationsversustreatment efforts to control older, previously-detected patches. Our new spatio-temporal economic model will examine the trade-offs between these objectives, and help to select a subset of survey sites to minimize the expected cost of damage and potential future damages due to undetected invaders under a limited budget. The new model will also incorporate the uncertainty about the invader's presence in the search region. Thus, our modeling and data collection efforts will first focus on determining the probability of infestation in each site, inspection and detection time, the cost of search per unit time, and the likelihood of detecting an infestation if it is present. Collaborator Dr. Greg Houseman has collected significant data from his field work. This data includes sericea germination rate, sericea seed production, stem density, sericea stem frequency over the landscape, and sericea survival rate from multiple pastures in Kansas (e.g., Brink pasture, Vestring Pasture, Youngmeyer Pasture, Stotts pasture, Browning pasture). Next step of our research involves analyzing this big field-based data set, and using it for parameter calibration and model validation for the search and control model. Additionally, we will continue the research to address the challenges in the solution of the highly complex invasive species search and control problem and develop specialized algorithms for solving the problem. Furthermore,the research teamwill continue the effortsto disseminate research findings through leading conferences and invited seminars.
Impacts
What was accomplished under these goals?
Goal A 1) The primary empirical effort in the second year of the experiment was data collection from 53 plots established on 7 ranches within the Flint Hills of Kansas. Over 1800 individual plants were marked on these sites in 2016. We rechecked each individual in 2017 and, if dead, a replacement was marked for further quantification of survivorship. Preliminary results suggest that soil fertility had little effect on survivorship. Sericea seed production was highly variable in the first year examined.Projected population growth rates were also highly variable substantiating the need for examining multiple sites before developing control recommendations. 2) After one year of herbicide application, there was no evidence that soil fertility influenced effectiveness of herbicide on Sericea. Goal B Under Goal B Objective 3, the CO-PD Buyuktahtakin with the research team have developed a novel optimal search and treatment model for effectively controlling invasive species. In particular, a new search-path optimization model, which takes into account the realistic nature of the detection and treatment processes as well as budget restrictions, has been developed. The results of the model performance are then applied tothe invasion of sericea (Lespedeza cuneata). Empiricaldata related to the growth and spread dynamics of the plantwas used to generate real-world population scenarios with the help of mathematical simulation. In the next step, the optimization model was applied to a 10x10 grid landscape for fifteen-year time-period based on the simulated setting. In those experiments, we have used the past data available in the literature (e.g., Büyüktahtakin, E., Kibis, E. Y., Cobuloglu, H. I., Houseman, G. R., & Lampe, J. T., 2015) and possible dispersal scenarios from the simulation experiments.
Publications
|
Progress 03/01/16 to 02/28/17
Outputs
Target Audience:During this first year of the project, we have established a network of ranchers in eastern Kansas that have either become informed about the details of the project or partners in this new research. Additionally, we have made contact with land management professionals, and state and federal professionals that assist the ranching community in our region. Changes/Problems:One unexpected development in the first year of the project was that we were unable to find a sufficient number of ranchers using early season, double stocking cattle management. Consequently, we modified the design of the project to include several cattle management strategies such as 90-day stockers, 90-day cow-calf, year-round cow-calf. Although this will reduce statistical power within a specific cattle management, it will greatly improve the generality of the modeling effort, because we will have field data from a much broader set of cattle management approaches representative of the region. A second modification relates to the detection experiment. After the initiation of the project, we learned of a GPS-sprayer apparatus that would allow us to quantify the location and abundance of Sericea across whole landscapes. This is an exciting opportunity because it allows us to quantify the spatial distribution and spread on multiple ranches utilizing the ongoing effort to control Sericea through spot-spraying. We think this will greatly enhance the spatial modeling of Sericea. Additionally, we are using this approach to quantify Sericea detection rather than the plot method included in the proposal. Ranchers will use dye during spot spraying so that we can survey randomly selected points to quantify the number of plants detected (sprayed). What opportunities for training and professional development has the project provided?The establishment of the empirical work relied heavily on assistance from graduate and undergraduate students. Under the supervision of the PDs, students received training in ranching practices, weed biology, and general principles of research design. For the modeling component, graduate students have received training in the use of optimization modeling and analysis with applications to invasive species. How have the results been disseminated to communities of interest?The PD was invited to give a presentation of the preliminary results to the Tallgrass Legacy Alliance. This group consists primarily of ranchers along with representatives from state and federal agencies and is keenly interested in the results of the project. The Co-PD and her students has organized and chaired the Energy, Natural Resources and the Environment (ENRE) sponsored session, titled "Optimal Surveillance and Control of Bio-Invasions" in the 2016 INFORMS Annual Meeting in Nashville, TN and presented results and findings of this research at this ENRE session. The Co-PD has been invited to present her research on "Multi-Stage Stochastic Optimization Approaches for Surveillance and Control of Invasive Species" at the INFORMS MIF Rising Young Scholars Award Session at the 2016 INFORMS Annual Meeting (The CO-PD is honored to receive the INFORMS MIF Early Career Award by the INFORMS Society.) The Co-PD has chaired a session on "Logistics" at the 2016 INFORMS Optimization Society Conference that was held in Philadelphia, PA. She presented her research on dynamic optimization and approximation algorithms in this conference. The Co-PD and her students who worked on this project have presented technical papers for invasive species control and optimization research at the 2016 INFORMS Annual Meeting and at the leading conference in the institute of industrial and system engineering (IISE), namely the 2016 IISE Annual industrial and system engineering research conference (ISERC). The Co-PD gave a seminar on "Problem Solving & Decision Making" to middle school students at a STEM camp in Wichita. She also had a brief presentation about "Industrial and Systems Engineering and Applications to Invasive Species Management" at the Wallace Invitational for Scholarships in Engineering (WISE) Panel organized by the Society of Women Engineers (SWE) and College of Engineering at Wichita State University. What do you plan to do during the next reporting period to accomplish the goals?In the next year, PD Houseman and his team plans to continue data collection necessary to refine the bioeconomic model and quantify the demography of the species. He also plans to refine the approach for quantifying the probability of Sericea detection. Co-PD plans to develop the search and control model in Goal B Objective 3 using the 2-year data provided by the PD Houseman. Additionally, she will continue working on specialized algorithms for solving the highly complex search and control problem inherent to invasive species control.
Impacts
What was accomplished under these goals?
Goal A The primary empirical effort in the first year was the establishment of research sites and collection of initial data. Led by PD Houseman, we established fifty-three, one-acre experimental plots distributed across eight ranches that represent different soil conditions and grazing management practices. We began data collection on Sericea density and survivorship for four life stages that will be critical components for refining the bioeconomic model. Additionally, we marked individuals that will allow us to accurately address the effectiveness of herbicide on the different life stages. Goal B Co-PD Buyuktahtakin developed a preliminary mathematical modeling framework for the search & control model that utilizes a general approach without reliance on the actual field data. This framework will be adapted to studying Sericea with the first and second year field data as they become available. Additionally, the co-PD has worked on developing general-purpose optimization algorithms that would be applied to solve the search and control model.
Publications
|
|