Progress 01/01/24 to 12/31/24
Outputs
Target Audience:University Students in Engineering and Biological Sciences. Elementary School Students. Ecotoxicology practitioners. Aquaculture practitioners. Cyber-Physical Systems practitioners. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?One of the doctoral students participating in the projectdefended his doctoral dissertation at UNT in June 2024 and is currently a Postdoctoral Fellow at the University of British Columbia. The dissertation citation is given under Products. One doctoral student in electrical engineering is supported with the grant to work on the deep learning based video monitoring and analysis. One doctoral student in environmental scienceis supported with the grant to work on the recirculating aquaculture systems. How have the results been disseminated to communities of interest?A poster presentation was showcased at the annual North American Society of Toxicology and Chemistry (SETAC) meeting, held October 20-24, 2024 in Ft. Worth, TX. This event brought together thousands of professionals and students from academia, industry, and government sectors worldwide. The poster informed this community of the acute toxicity of ammonia to juvenile M. rosenbergii and the feasibility of implementing real-time respirometry as an early detection method for detecting harmful changes in water quality. https://www.setac.org/discover-events/global-meetings/setac-north-america-45th-annual-meeting/program/daily-schedule.html. Please see list of products. A similar poster was also presented at the Texas Chapter of the American Fisheries Society (TCAFS) Annual Meeting, held February 22-24th in Nacagdoches, TX. This event brought together hundreds of professionals and students from academia, industry, and government sectors across Texas. https://units.fisheries.org/tx/tc-meetings/2024-tcafs-meeting-home-page/. Please see list of products. We participated in the NSF Cyber-Physical Systems (CPS) Principal Investigator's Meeting 2024 in Nashville, TN, and disseminated our project results by presenting a poster (cited in the list of Products section) and posting a recordedlightning talk.This meeting is attended by PI's of all NSF active CPS projects, and thus our project became visible as an important contribution of CPS in Agriculture. What do you plan to do during the next reporting period to accomplish the goals?Now that we have verified that respirometry can be used to detect subtle, sublethal stress responses to changes in water quality, our next step will be to design and implement custom-made respirometers for prawns that can be integrated in-line with our current RAS, test and optimize these respirometry systems, and begin integrating with machine learning. We will submita manuscript for review to the journal, Aquaculture, to disseminate our results on ammonia toxicity and the feasibility of using respirometry as an early detection system to the wider scientific community. Due to the low survival results obtained we have opted to try another species (Litopenaeus vannamei) that can handle higher densities in both nursery and grow out to develop a growth and survival model using brine from desalination plants and comparing it with artificial seawater at the optimum salinity for the species (20 ppt). Now that we have a tool to effectively monitor weekly growth and survival of the animals without having to sacrifice the animals and interfere with the densities as well as a species that can handle high densities both in nursery ang grow out, we will perform a comparative study throughout the full grow out cycle (1g-20g), calibrate and refine the growth and survival models for artificial seawater and brine from desalination plants. For the deep learning part of the project, we will continue to expand the models that we have developed so far to include the capabilities to monitor and estimate the growth and behavioral patterns of prawns under stress conditions. We will also further experiment with data collection, data labeling, and training deep learning models by incorporating video monitoring capability in the respirometry system. We will be submitting apaper to IEEE Sensors Journal to disseminate our results on size-adaptive density map for prawn postlarvae counting. We also plan to implement the nutrient separation component of the project to contribute to Objective 2, as well as develop the model of oragnism growth as part of Objective 4.
Impacts
What was accomplished under these goals?
With respect to objective 3, we completed our experiment designed to assess whether metabolic (i.e., oxygen consumption) responses could be used as an early, sublethal indicator of stress to ambient ammonia accumulation for juvenile prawns. To this end, we tested how increasing waterborne ammonia concentrations affect routine metabolic rate (RMR) of juvenile Macrobrachiumrosenbergii using static intermittent respirometry (n=6-9/treatment). The step-wise ammonia concentration gradient tested in the experiment proceeded as follows: 0, 3, 6, 9, 12, 15, 18, 21, 24, 30, 36, 42, and 48 mg/L total ammonia nitrogen (TAN). To evaluate potential interactive effects of ammonia responses with salinity, we conducted these experiments at 3 salinities: 1, 5 and 10 ppt. Results revealed a significant interaction between salinity and ammonia, driven primarily by the response at 10 ppt, alongside a significant main effect of ammonia, where RMR increased with higher ammonia concentrations. However, no significant effect of salinity was observed across trials. Importantly, these findings demonstrate the feasibility of implementing a sentinel respirometry system for detecting changes in water quality in real-time by monitoring metabolic rates. In terms of outcomes, this metabolic information will provide an early warning system for aquaculturalists to prevent episodes of mortality or lack of growth due to chemical stressors. Additionally, the aquarium display was implemented at the Elm Fork Education Center as a contribution to Objective 5. With respect to objective 4, we have completed a full grow out cycle in the Recirculating Aquaculture Systems (RAS) starting with 1.5g juveniles of M. rosenbergii. Even though growth was satisfactory reaching 30g in 4 months, survival of M. rosenbergii in the recirculating aquaculture system was not satisfactory (<10%) due to the highly cannibalistic nature of the species. Ammonia and nitrite were kept at very low levels (<1ppm) while nitrate levels reached 150 ppm, showing an appropriate functioning of the biofiltration system. Less than 1% water exchange/day was needed to keep the Nitrate levels below 150 ppm and compensate for evaporation.In order to measure weekly growth and survival without having to sacrifice the animals, we have developed an image recognition model for counting, measuring and estimating biomass of the animals. We have been developing the video monitoring system and deep learning models for computer vision-based aquatic behavioral assessment and toxicity prediction, which contribute to the objectives 1 and 3. We have set up a video monitoring system in the lab and we have been collecting data using the system for developing deep learning models. For deep learning computer vision-based analysis, we have developed a novel deep-learning framework that integrates size-adaptive density map estimation with a multiscale feature extraction network for precise and robust prawn postlarvae counting. The size-adaptive density map dynamically adjusts to varying postlarvae sizes, ensuring accurate counting even in dense and heterogeneous environments. Meanwhile, the multiscale feature extraction network captures features across multiple resolutions, effectively detecting tiny and overlapping postlarvae. Experimental evaluations on a newly curated prawn postlarvae dataset demonstrate that P2CountNet significantly outperforms state-of-the-art counting methods, achieving lower mean absolute error (MAE) and mean squared error (MSE). Notably, the proposed method achieves an average counting accuracy of 94.32% for datasets with approximately 1,000 postlarvae. The datasets used in this research are collected with the video monitoring system that we have deployed in the lab.
Publications
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2024
Citation:
Emadi, C., Dos Santos Neto, F., Bohenek, J., Smithers, B., Acevedo, M. F., Mager, E., 2024 Effects of Salinity on the Toxicity and Real-Time Metabolic Rate Responses of Acute Ammonia Exposure to Juvenile Macrobrachium rosenbergii. Society of Environmental Toxicology and Chemistry North America Annual Meeting, Poster, Fort Worth, TX. https://www.setac.org/discover-events/global-meetings/setac-north-america-45th-annual-meeting/program/daily-schedule.html
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2024
Citation:
Acevedo, M.F., Li, X., Mager, E. M., Smithers, B., 2024 Integrating sensors, controls, and ecotoxicology with decoupled aquaponics using brackish groundwater and desalination concentrate for sustainable food production, NSF Cyber-Physical Systems Principal Investigators' Meeting, Poster, National Science Foundation, Nashville, TN.
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2024
Citation:
Emadi, C., Dos Santos Neto, F., Bohenek, J., Smithers, B., Acevedo, M. F., Mager, E., 2024. Toxicity Assessment and Real-Time Metabolic Trait Responses of Juvenile Macrobrachium rosenbergii to Ammonia Exposure at Different Salinities, Texas Chapter of the American Fisheries Society Annual Meeting, Poster, Texas Chapter of the American Fisheries Society, Nacogdoches, TX. https://units.fisheries.org/tx/wp-content/uploads/sites/19/2025/01/TCAFS-2025-Program_Updated.pdf
- Type:
Conference Papers and Presentations
Status:
Published
Year Published:
2024
Citation:
Neto, F., Dang, T.H., Smithers, B., Li X., Mager, E., Acevedo, M.F.. 2024, Image Recognition Model for Crustacean Postlarvae Counting and Size Estimation. Poster presentation UNT Research Day, Denton, TX.
- Type:
Theses/Dissertations
Status:
Published
Year Published:
2024
Citation:
Emadi, C. (2024) Physiological Impacts of Anthropogenic-Induced Stressors on Freshwater Animals. [Doctoral Dissertation, Department of Biological Sciences, University of North Texas]. https://doi.org/10.12794/metadc2356211
|
Progress 01/01/23 to 12/31/23
Outputs
Target Audience:Ecotoxicology, aquaculture industry, CPS community, and desalination community. Changes/Problems:
Nothing Reported
What opportunities for training and professional development has the project provided?The implementation and operation of the RAShas provided opportunities for training and professional development for undergraduate and graduate students in the field of biology, engineering, and environmental sciences. In addition to providing opportunities in their own disciplines, the project offered a unique platform for interdisciplinary training. For example, biology students have learned about electronic monitoring and networking, while electrical engineering students have learned how to work with aquatic systems. This project is engaging doctoral students in both biology andelectrical engineering by workingon novel respirometry assessments, comprehensive integration of animal life cycle into aquaculture systems, advancedvideo monitoring system development, and computer vision-based aquatic behavioral assessment. Development of the real-time water quality monitoring system was accomplished by a team of senior design electrical engineering students, therefore having animpact on the education and training of undergraduate engineering students, who have since graduated and are now part of the workforce. How have the results been disseminated to communities of interest?SETAC North America: A poster presentation was showcased at the annual North American Society of Toxicology and Chemistry (SETAC) meeting, held November 12-16, 2023 in Louisville Kentucky. This event brought together hundreds of professionals from academia, industry, and government sectors worldwide. The poster informed this community of the importance of managing the concentrate from desalination processesand the challenges of developing brackish water aquaculture.https://www.setac.org/discover-events/global-meetings/setac-north-america-44th-annual-meeting/program/daily-schedule.html Please see list of products: Emadi, C., F. Dos Santos Neto, B. Smithers, M. Acevedo, and E. Mager. (2023) Toxicity Assessment and Real-Time Metabolic Trait Responses of Juvenile Macrobrachium rosenbergii to Ammonia Exposure at Different Salinities. Society of Environmental Toxicology and Chemistry North America Annual Meeting, Louisville, KY. 7th Integrative Biology Workshop: A poster was presented at this workshop, hosted from October 27-31, 2023 in Toluca, Mexico. This workshop marked the 7th meeting in a series of ongoing collaborations between the University of North Texas and Autonomous University of the State of Mexico in Toluca, Mexico State. With a focus on Integrative Biology, this event primarily draws participation of students and faculty from the two universities, thus promotinginternational collaboration in the topics of this project. See workshop schedule at the following linkhttps://acrobat.adobe.com/link/review?uri=urn:aaid:scds:US:1573d044-7533-3bce-a1c7-dd924c2cf803 Please see list of products: Emadi, C., F. Dos Santos Neto, B. Smithers, M. Acevedo, and E. Mager. (2023) Toxicity Assessment and Real-Time Metabolic Trait Responses of Juvenile Macrobrachium rosenbergii to Ammonia Exposure at Different Salinities. 7th Integrative Biology Workshop, Toluca, Mexico. Water Innovation and Networking Workshop: We gave a platform presentation at the 7th Annual Water Treatment Innovation and Networking (WIN) Workshop, held on October 23rd and 24th, 2023 in Alamogordo, NM. It was hosted by the Brackish Groundwater National Desalination Research Facility (BGNDRF) and provided a platform for clients to share work in desalination and impaired water treatment. Industry leaders, researchers, and experts from around the world convened to explore innovations and foster collaboration. Link: https://www.usbr.gov/research/bgndrf/win.html Please see list of products: Emadi, C., F. Dos Santos Neto, E. Mager, B. Smithers, X. Li, and M. Acevedo. (2023) Desalination Concentrate as a Potential Resource for Inland Aquaculture. U.S. Bureau of Reclamation, Annual Water Innovations and Networking Workshop, Alamogordo, NM. What do you plan to do during the next reporting period to accomplish the goals?During the next reporting period, we are planning to 1) finalize the hardware and software configurations of the video monitoring system for effective operation integrated with the RAS, water quality monitoring system, and the respirometry assessments,2) using deep learning models, develop amethod to determineunique behavioral patterns of prawns under stress conditions,3) experiment with data collection, data labeling, and training deep learning models by incorporating video monitoring capability in the respirometry system, 4) Once thecontrol respirometry experiments at all salinities are analyzed, we will completeammonia spiking experiments at all salinities, and integrate the findings with the bioassay results, 5) design and implement custom-made respirometers for two life stages/sizes of prawns that can be integrated in-line with our current RAS, 6) test and optimize these respirometry systems,and begin integrating with machine learning to serve as a sentinel early-detection system for changes in water quality, 7) develop models of animal growth and survival, and use the experimental results of respirometry and bioassays for callibration and validation, 8) develop the nutrient separation system to improve water quality in the RAS, while providing a nutrient rich stream to the future hydroponics system,9) implement the aquarium display and lesson plans for outreach activities at the Elm Fork Education Center so that it will be available for all visiting school districts in North Texas area, 10)provideopportunities to underrepresented groupsby planning and conducting summer research experiences for high school and undergraduate students, and 11) continue to disseminate project results by presenting at conferences and submitting manuscripts to peer-reviewed journals.
Impacts
What was accomplished under these goals?
In order to accomplish all objectives above, three Recirculating Aquaculture Systems (RAS)were implemented in August 2023, each consisting of independent recirculating loops comprised of one 25-gallon tank and one 5-gallon biofilter. Each RAS has a flowrate of ~0.5 gallon per minute, which corresponds to ~30 turnovers per day. The systems have been operating continuously since then (with less than 1% water exchange/day) and are being tested with Macrobrachium rosenbergii (Giant river prawn)post-larvae (under 1g), juveniles (1-5g), and adults (>5g). The animals that were stocked in August 2023 have grown and matured in the system. Once the efficacy of the RAS system has been tested with success on all life stages of the animal, we will start to conduct trials using brine from desalination plants, monitor water quality parameters, compare survival and growth in different life stages, and start the implementation of the in-line respirometer. Working towards objective 1, four water quality parameters: pH, dissolved oxygen (DO), electrical conductivity (EC), and temperature are monitored continuously in the threeRAS mentioned above. This was accomplished by developinga real-time monitoring system comprised of a sensor network and a single board computer that compiles the data totransmitto a webserver anddatabase system.Using a human machine interface (HMI), these readings are available to the biology students that work with the animals on a daily basis, thereby providinginstant information onthequality of the water while they assess the health and growth of the animals. This setup serves as a pilot for what the operators of the future testbed can rely upon for itssuccessful operation. Therefore, this technologywill have an impact on the tools that the aquaculture community can use to improve the operation of their systems. The single board computer mentioned above is aRaspberry Pi (RPi), which is a low-cost and small footprint device that can be programmed with public domain software. This RPiis the brain of themonitoring system andexecutesa Graphical User Interface (GUI) created and powered through Node-Red for real-time data visualization, and data storageinto influxDB database. An advantage ofthis design is that the system can be scaled up to a larger number of RAS and to other water quality parameters. Indeed, each sensor operates independently, allowing for the system to be easily expanded or modified with additional sensors as needed. This design will have an impact on the aquaculture community because it can be adapted to different sizes of RAS implementations or to meet varying requirements of specific aquatic organisms. We have been developing thevideo monitoring system and deep learning models for computer vision-based aquatic behavioral assessment and toxicity prediction, which contribute to the objectives 1 and 3. The video monitoring system includes a RPias the field computing platform, a USB web cam, a webserver running in the RPi, web services designed for on-demand video streaming, and video sink softwarefordata collection. We areexperimenting with different configuration approaches for more effective monitoring and data collection from aquaponic systems. For the deep learning part of the project,we have started training YOLO v8 models with the video data samples collected from the animalscultured in the threeRAS mentioned above. We have been able to successfully detect and track animalsin the video stream data using the trained YOLO models. Currently, we are in the process of refining the training process and analyzing prawnsbehavioral patterns from the detection results obtained using the trained models, which is an essential step to progress towards behavioral assessment-based toxicity prediction. Similarly to the water quality monitoring system, successful demonstration of the validity of this approach will provide significant enhancement of tools that theaquaculture community can use make sure health and growth of the cultivated organisms. With respect to objective 3, as a first step toward integrating respirometry into a real-time toxicological assessment of water quality, we performed 48-hour bioassays at three salinities (1, 5 and 10 ppt) and a standard pH of 8.2 using ~40 day-old post-larvae exposed to a range of total ammonia concentrations as a focal toxicant.We selected ammonia because of its high relevance in high-density aquaculture systems given its propensity to accumulate, and thereby negatively affect the survival, growth, and overall health of aquatic animals. Survival data was collected, and corresponding median lethal (LC50) concentrations calculated, for four timepoints at each salinity: 3, 6, 24 and 48 hours.From this data, a concentration was selected (20 mg/L total ammonia) to begin testing whether ammonia spikes could be detected using real-time respirometry. To this end, we first collected control baseline data using juvenile shrimp in individual static respirometers at each of the three salinities (n=8/salinity). These tests were designed to not only inform baseline metabolic data, such as routine and standard metabolic rates (RMR and SMR, respectively), but also to inform as to the time required for the animals to habituate to the chambers (i.e., the initial period required to achieve RMR).Knowledge of the habituation period is important to ensure that the time of spiking with ammonia occurs after habituation to maximize sensitivity in the measured response.Next, we initiated a limited number of tests (n=2) at the intermediate (5 ppt) salinity to test whether a change in metabolic rate in response to the ammonia spike could be detected.While more tests and statistical analyses are pending, our preliminary predictionfrom inspection of the data, is that the observed prolonged increase in metabolic rate immediately following the ammonia spike is likely significant, thusproviding initial support that the approach will have success. In addition to its contribution to meetobjective 1, this information provides the basis to inform the development of models of organism growth and survival, which are the core of objective 4. In terms of outcomes, this metabolic information will provide an early warning system for aquaculturalists to provent episodes of mortality or lack of growth due to chemical stressors.
Publications
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2023
Citation:
Emadi, C., F. Dos Santos Neto, B. Smithers, M. Acevedo, and E. Mager. (2023) Toxicity Assessment and Real-Time Metabolic Trait Responses of Juvenile Macrobrachium rosenbergii to Ammonia Exposure at Different Salinities. Society of Environmental Toxicology and Chemistry North America Annual Meeting, Louisville, KY.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2023
Citation:
Emadi, C., F. Dos Santos Neto, B. Smithers, M. Acevedo, and E. Mager. (2023) Toxicity Assessment and Real-Time Metabolic Trait Responses of Juvenile Macrobrachium rosenbergii to Ammonia Exposure at Different Salinities. 7th Integrative Biology Workshop, Toluca, Mexico.
- Type:
Conference Papers and Presentations
Status:
Accepted
Year Published:
2023
Citation:
Emadi, C., F. Dos Santos Neto, E. Mager, B. Smithers, X. Li, and M. Acevedo. (2023) Desalination Concentrate as a Potential Resource for Inland Aquaculture. U.S. Bureau of Reclamation, Annual Water Innovations and Networking Workshop, Alamogordo, NM.
|