Duckweed Final Report

0

Evaluating the Feasibility and Effectiveness of Duckweed Phytoremediation in Lake Sevan

An Interactive Qualifying Project Report submitted to the Faculty of WORCESTER POLYTECHNIC INSTITUTE in partial fulfillment of the requirements for the degree of Bachelor of Science

By: Eric Bormann Phillip Durgin Alexander Gladu

Date: May 12th, 2021

Report Submitted to: Garabet Kazanjian, AUA Acopian Center for the Environment Alen Amirkhanian, AUA Acopian Center for the Environment

Advisors: Aaron Sakulich, Worcester Polytechnic Institute Norayr Ben Ohanian, American University of Armenia

This report represents the work of WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review. For more information about the projects program at WPI, please see http://www.wpi.edu/academics/ugradstudies/project-learning.html 1

Table of Contents

Abstract 3

Executive Summary 4 Project Goal and Objectives 4 Results of Geographic and Pollution Data Analysis 5 Results of the Assessment of Local Attitudes 6 Results of the Opportunity Analysis 6 Results of the Pilot Farm Development 7 Recommendations 8 Farm Location Recommendations 8 Duckweed Post-Harvest Application Recommendations 9 Farm Design Recommendations 9

Acknowledgments 10

Authorship 11

List of Figures 12

List of Tables 13

Introduction 14 History and Importance of Lake Sevan 17 Table 1. Species Endemic to Lake Sevan 19 Eutrophication, Metal Pollutants and Duckweed Water Treatment 20 Challenges with Maintaining a Duckweed Farm 24 About the AUA Acopian Center for the Environment 27 Summary 28

Methodology 29 Objective 1. Geographic and Pollution Data Analysis 30 Table 2. NSF Water Quality Index Parameter Weights 31 Objective 2. Assessment of Local Knowledge and Attitudes 34 Objective 3. Opportunity Analysis of Duckweed Farming 36 Objective 4. Design of a Duckweed Pilot Farm 41

Results 44 Objective 1. Geographic and Pollution Data Analysis 45 Table 3. Water Quality and Nutrient Concentration Rankings 45 Table 4: Municipality Population Data Circa 2011 47 Table 5: Hydrology Data Calculated for the Gavaraget, Martuni, and Masrik Rivers 47 Objective 2. Assessment of Local Attitudes 48 2

Objective 3. Opportunity Analysis of Duckweed Farming 55 Table 6. Biomass Estimations for the Months Between March and October 56 Objective 4. Design of a Duckweed Pilot Farm 61

Recommendations 67 Farm Location Recommendation 67 Duckweed Post-Harvest Application Recommendation 67 Farm Design Recommendations 72

Conclusions 74

References 76

Appendix A - International Reports of the Ministry of Nature Protection with Wetland Measurements 84

Appendix B - Guiding Questions for Interviews with Municipality Representatives 85

Appendix C - Algorithm Variable Information 86

Appendix D - Ethanol Production Calculations 87 3

Abstract

Recurring events of eutrophication, caused by reduced water levels and increased nutrient concentrations, have damaged the Lake Sevan ecosystem. The American University of Armenia’s

Acopian Center for the Environment is seeking to use duckweed as an effective, low-cost, and economically beneficial method of phytoremediation. The team identified locations most in need of water treatment, researched the feasibility of post-harvest applications, and determined local attitudes towards the project. Using this information, a pilot farm was designed for future implementation and evaluation. 4

Executive Summary

Facing the collapse of its ecosystem due to pollution, Lake Sevan in Armenia is yet another tragic example of the consequences of human exploitation and degradation of natural resources. Our project, sponsored by the American University of Armenia (AUA) Acopian Center for the Environment, was designed to help establish possible approaches that could limit further nutrient and pollutant inflow into

Lake Sevan from its tributaries. By doing so, further progress can be made to maintain and restore the fragile ecosystem surrounding the lake through sustainable and economically beneficial methods. The method of phytoremediation which was examined involved the usage of duckweed, a floating macrophyte, to uptake high concentrations of phosphorus and nitrogen from the mouths of rivers flowing into the southern body of Lake Sevan. Previous research has been conducted into the feasibility of using duckweed as a water treatment system. Alongside being able to remove excess nutrients from wastewater, duckweed can remove heavy metals and has many post-harvest applications, making it an ideal plant for phytoremediation systems.

Project Goal and Objectives

The goal of this project was to determine the feasibility and impact of implementing a duckweed-based phytoremediation system in the tributaries surrounding Lake Sevan. To accomplish this task, objectives were developed in order to effectively work towards the completion of this goal over the duration of the project. An outline of our objectives is displayed in the figure below. 5

1. Objective 1 sought to analyze previously collected water quality data in order to determine rivers

able to support duckweed growth with the highest contributions to the eutrophication process.

2. Objective 2 focused on the collaboration with the municipalities surrounding the locations which

were best suited for implementation. Interviews were conducted to determine the attitudes

towards duckweed farm implementation, as well as to reveal logistical challenges and local

concerns.

3. Objective 3 estimated the economic impact duckweed farming would create through its various

post-harvest applications in order to incentivize continued collaboration with local municipalities.

4. Objective 4 looked to design a pilot farm with the goal of future implementation. This would

allow the Acopian Center to corroborate our research and initial estimations with pilot testing.

Results of Geographic and Pollution Data Analysis

The work we conducted for our first objective used nutrient concentration and water quality data previously collected for a report conducted under the European Union Water Initiative Plus (EUWI+).

Rivers were ranked based upon their phosphate and nitrate concentrations, their ability to support optimal duckweed growth, and their National Sanitation Foundation Water Quality Index score. Through the analysis of the initial three rankings, it was determined that the mouths of the Gavaraget, Masrik, and

Martuni rivers were locations most in need of water treatment while still able to harbor and optimize duckweed biomass yield. After ranking the ten rivers and selecting the Gavaraget, Masrik, and Martuni as rivers of interest, Google Maps was used to make a list of various towns and villages situated around the rivers. Eighteen towns were selected for conducting potential interviews. It was also determined through the analysis of hydrology data that the Gavaraget river has the largest contribution of nitrogen and phosphorus, 179,662 kg (396,087 lbs) and 29,619 kg (65,299 lbs) respectively, flowing into Lake Sevan. 6

Results of the Assessment of Local Attitudes

To better understand local attitudes towards the implementation of a duckweed phytoremediation system and estimate the support and infrastructure available for farm implementation, representatives from the previously identified municipalities surrounding Gavaraget, Masrik, and Martuni were interviewed. The general consensus was that although most of the communities supported the goal of the project, they felt that there was further testing that needed to take place before their municipalities would be willing to support project efforts. Further information regarding the effectiveness of post-harvest applications and the resources needed for farm development and implementation will need to be provided.

Results of the Opportunity Analysis

To determine the potential economic benefit of duckweed harvesting, an algorithm was developed to estimate the biomass yield from duckweed farming. This model used a variety of parameters that affect duckweed growth as inputs, including water temperature, photoperiod, pH, and nutrient concentration, in order to apply the procedure in a variety of environments. It was estimated that a total of 1,093 kg/m²

(223.864 lbs/ft²) of dry duckweed biomass could be harvested from the Gavaraget River. For the Martuni and Masrik rivers, this biomass value was estimated to be 755 kg/m² (154.636 lbs/ft²) and 774 kg/m²

(158.528 lbs/ft²) respectively.

After estimating biomass production, research shifted towards duckweed applications. Studies have shown that animals fed with duckweed supplementing comercial feed have a significant increase in size, likely a result of dried duckweed having a high protein content and a low fiber content. While this research has shown that duckweed has several benefits when used as feed for fish and other animals, the amount of duckweed used as feed and when the duckweed is fed to the animals varies depending on the animal. Looking first at poultry, various studies note that chicks have stunted growth when fed duckweed, but older birds were seen to have higher growth rates than their counterparts fed with commercial feed.

Despite several studies noting negative effects when fish are fed too much duckweed, the amount of 7 duckweed able to be fed to fish without seeing negative impacts varies study to study and seems to be dependent in part on the species of fish in the studies.

Research has also shown duckweed to be a promising standalone fertilizer and fertilizer amendment. Experiments conducted at Pennsylvania State University found that duckweed-based fertilizer is likely to have less nitrogen runoff compared to commercial fertilizer. In addition, duckweed increased the nutritional value of sorghum grown with duckweed as fertilizer. Research also shows that duckweed helps soil retain other nutrients such as carbon and calcium. Overall, duckweed performed similarly to commercial fertilizer, providing better nutritional value for plants with the drawback of significantly impaired germination.

Bioethanol research revealed that using duckweed as a feed source for bioethanol production has many benefits. Duckweed not only has a higher growth rate and longer growth period than traditional field crops, but also has less environmental impacts as it requires limited agricultural inputs and helps to treat wastewater. The main method required to produce ethanol from a duckweed-based feed source is through enzymatic hydrolysis for fermentable sugar production. Through this process, it was estimated that a single duckweed farm could produce between 153 L (40.42 gallons) and 357.42 L (94.42 gallons) of ethanol.

Results of the Pilot Farm Development

Based on the minimum growth temperature and optimal levels of starch and protein composition,

Lemna minor, Lemna gibba, and Lemna trisulca were selected for use in the duckweed pilot farms. The building of the farm will be dependent on the resources available, but will need to meet general specifications regardless. The shape of the pond should be narrow and long with a width under 6 m (20 ft) in order to reduce difficulty when harvesting. The depth of the pond should be between 1 to 1.5 m (3 to 5 ft) to ensure this as well. Ideally, the pond will be installed on flat land to limit the resources required for installation while the water should be still and in an area with minimal wind exposure. Systems must be in 8 place to ensure a water of slow current enters the pond and duckweed stays within the farm boundaries during water outflow. Considerations must also be made about who will manually be harvesting the duckweed as it requires a high degree of physical labor. Volunteers will be needed to harvest the duckweed and perform routine maintenance. Automated harvesting techniques, unlike manual methods, have very high initial investment costs in exchange for faster harvesting capabilities and minimal human involvement. These systems may be too costly or impractical for farms of this nature though. Finally, two methods of drying were researched: solar drying and oven drying. Solar drying is the cheapest drying method, only requiring black plastic sheets and sunlight, while oven drying is a much faster and costly process.

Recommendations

Based on the results of our research, we have consolidated recommendations into three topics for the Acopian Center to continue their efforts in designing a duckweed-based phytoremediation system:

1. Farm Location

2. Duckweed Post-Harvest Application

3. Farm Design

Farm Location Recommendations

We recommend that the Acopian Center deevelops a process to identify open land or water for pilot sites in the regions surrounding Gavaraget, Masrik, and Martuni. Land or water in regions in which there was initial support for the project, such as Karchaghbyur, should be considered primary locations for consideration. It is also important to consider implementation in regions surrounding rivers with the highest nutrient outflows, such as the Gavar region. 9

Duckweed Post-Harvest Application Recommendations

Although we were able to utilize an algorithm to estimate the growth rate and biomass output in conditions found in the Gavaraget, Masrik, and Martuni rivers, pilot farm testing is needed to corroborate our calculations. Based on the nature of the local economies, we recommend that harvested duckweed is used for animal feed supplements, compost, and fertilizer supplements. Since research into optimizing duckweed as animal and fish-feed research has varying conclusions, we recommend further research looking into the maximum amount of duckweed can be used to replace commercial feed before the negative impacts are seen in animals.

Farm Design Recommendations

In order to efficiently harvest the duckweed, the pond should be narrow. We recommend a width between 4.5-6 m (15-20 ft). Similarly, the depth of the pond should not be more than 5 ft deep (1.5 m).

The plot of land for the duckweed farm should be either inexpensive or flat enough that the initial construction cost is minimal. For the harvesting process we recommend that pond rakes are used as automated methods are too large an investment for communities with limited resources. Once the duckweed has been harvested, we recommend placing it onto black plastic sheets in the sun with the duckweed evenly spread across it. 10

Acknowledgments

We would like to express our gratitude to the following people for their contributions and support throughout the duration of this project. First and foremost, we would like to thank Garabet Kazanjian, our sponsor at the AUA Acopian Center for the Environment, for helping to direct our project and providing us with continual support. Without him we would have not been able to complete the project.

We would also like to acknowledge our advisors Norayr Ben Ohanian and Aaron Sakulich, for pushing us to go above and beyond while providing continual feedback and support throughout the duration of the project. We are also thankful for Carol Stimmel, whose guidance during ID 2050 laid the groundwork for our success during the duration of our work.

Our team would not be complete without Hovhannes Torosyan, our Armenia team member who provided useful insight and made interviews with municipalities possible. His work was a crucial part of our report and he helped to overcome the challenges presented by COVID-19.

Finally, we’d like to thank Alexander Arakelyan, an Armenian hydrologist for providing us with data and answering our questions regarding hydrology and water quality testing. Without this information, we would not have been able to conduct much of our research into nutrient concentration and farm location recommendation. 11

Authorship

Section Primary Author(s) Primary Editors(s)

Abstract Phillip Durgin N/A

Executive Summary Phillip Durgin N/A

Acknowledgements Phillip Durgin Eric Bormann

Introduction Phillip Durgin and Eric Bormann Alexander Gladu

Background Introduction Phillip Durgin Eric Bormann

History and Significance of Lake Sevan Phillip Durgin Eric Bormann

Eutrophication, Metal Pollutants and Alexander Gladu Eric Bormann and Phillip Durgin Duckweed Water Treatment

Challenges of Maintaining Duckweed Farms Eric Bormann Alexander Gladu and Phillip Durgin

About the AUA Acopian Center Eric Bormann and Phillip Durgin N/A

Background Summary Phillip Durgin Eric Bormann

Introduction of Methodology Eric Bormann Phillip Durgin

Geographic and Pollution Data Analysis Phillip Durgin Eric Bormann

Assessment of Local Attitudes Phillip Durgin Eric Bormann

Opportunity Analysis Phillip Durgin and Eric Bormann Alexander Gladu

Design of a Duckweed Pilot Farm Eric Bormann and Alexander Gladu Phillip Durgin

Results Introduction Phillip Durgin Eric Bormann

Geographic and Pollution Data Analysis Phillip Durgin and Eric Bormann N/A

Assessment of Local Attitudes Phillip Durgin and Eric Bormann N/A

Opportunity Analysis Phillip Durgin and Eric Bormann N/A

Design of a Duckweed Pilot Farm Eric Bormann and Alexander Gladu Phillip Durgin

Recommendations Eric Bormann, Phillip Durgin, and N/A Alexander Gladu

Conclusions Phillip Durgin and Eric Bormann N/A

Appendix A Phillip Durgin N/A

Appendix B Eric Bormann, Phillip Durgin, and N/A Alexander Gladu

Appendix C Phillip Durgin N/A

Appendix D Eric Bormann N/A 12

List of Figures

Name Page

Figure 1. Physical Map of Armenia 17

Figure 2. Reduction of Lake Sevan Water Levels 18

Figure 3. A Cluster of Fronds of the Species Lemna minor 22

Figure 4. A Duckweed Pond in Canada 23

Figure 5: Satellite Image of Blue-Green Algae Bloom Occuring on 44 July 20th, 2020

Figure 6. The Locations of the Gavaraget, Masrik, and Martuni River 46 Mouths

Figure 7: Local Fishery in the Sarukhan Village, Gavar Region 48

Figure 8: Local Ecotourism Pond, Artsvaqar Village, Gavar Region 51

Figure 9: Karchaghbyur River near the same named village, Vardenis 53 Region

Figure 10: Equations Used for Biomass Algorithm Calculations 55

Figure 11: Blueprints of the Least Expensive Duckweed Pond Design 64

Figure 12. How a PondSkim Would Be Used to Collect Duckweed 65 Floating on the Surface of the Water

Figure 13: Chlorosis Visible in the Leaves of a Strawberry Plant 71 13

List of Tables

Name Page

Table 1: Species Endemic to Lake Sevan 19

Table 2: NSF Water Quality Index Parameter Weights 31

Table 3: Water Quality and Nutrient Concentration Rankings 45

Table 4; Municipality Population Data Circa 2011 47

Table 5: Hydrology Data Calculated for the Gavaraget, Martuni, and 47 Masrik Rivers

Table 6: Biomass Estimations for the Months Between March and 56 October 14

Introduction

Facing the collapse of its ecosystem due to pollution, Lake Sevan in Armenia is yet another tragic example of the consequences of human exploitation and degradation of natural resources. Our project, sponsored by the American University of Armenia (AUA) Acopian Center for the Environment, was designed to help establish possible approaches that could limit further nutrient and pollutant inflow into

Lake Sevan from its tributaries. By doing so, further progress can be made to maintain and restore the fragile ecosystem surrounding the lake through sustainable and economically beneficial methods.

For centuries, Lake Sevan has provided Armenia with a source of hydropower, irrigation, and fishing grounds, as well as a location for recreation and tourism opportunities. Despite the great demand placed upon its resources, the lake supplied the Armenian people without major problems until the 1930s.

Around this time, the Armenian Soviet Socialist Republic began to use Lake Sevan to irrigate the Ararat

Valley and generate hydroelectric power. This was done by increasing the outflow of water in the Hrazdan

River, Lake Sevan’s only distributary. The changes the Soviet Union made to the Hrazdan River increased the outflow of water from the lake by approximately 1,400%, leading to a 20 m (65 ft) decrease in water level, increasing erosion and sedimentation rates of the exposed lakebed (Lind & Taslakyan, 2005). After overcoming the energy crisis that plagued the country since its independence in 1991, the Armenian government has taken steps to raise the water in the lake to help restore and preserve its ecosystem.

Despite these efforts, nitrate and phosphate levels in the water have gradually been increasing. These nutrients are introduced to the lake along with other pollutants from agricultural runoff, untreated wastewater, and increased sedimentation rates. With the increasing nitrate and phosphate levels, and several heavy metal concentrations exceeding water quality standards, Lake Sevan needs water treatment.

Previous research has been conducted into the feasibility of using duckweed, a floating macrophyte, as a water treatment system. Macrophytes are aquatic plants that can be seen with the naked eye and consist of either entirely floating or partially submerged parts (Dhote et al., 2009). Floating macrophytes are ideal for phytoremediation and waste treatment due to their high reproduction rate, 15 which increases nutrient removal. They are also much easier to harvest in comparison to their submerged counterparts. Furthermore, duckweed is cold-tolerant and has been shown to reduce the growth of damaging algae. Duckweed has been extensively studied for its use as a primary and secondary waste treatment for fish culturing ponds, domestic sewage, and dairy waste lagoons. These studies show that duckweed can remove most of the nitrogen and phosphorus present from sewage treatment ponds, leaving effluent behind that can be used in agricultural irrigation. Alongside being able to remove excess nutrients from wastewater, duckweed can remove heavy metals, making it an ideal plant for phytoremediation systems.

In search of an effective yet low-cost method of phytoremediation to remedy this issue, the AUA

Acopian Center for the Environment is interested in the efficacy of implementing duckweed farms in the tributaries of Lake Sevan. Before this can occur, considerations regarding the feasibility of duckweed farming in this type of environment must be considered. Further research into the type and concentration of nutrients present in Lake Sevan’s water, the most suitable locations for farm development, as well as the ability of these farms to be maintained are just some of the variables that need to be considered before further decisions can be made on how to progress. It is also important to understand ways in which duckweed farming can provide further opportunity for surrounding communities and incentivize continued collaboration between community members and the AUA Acopian Center for the Environment.

Research into the possibility of using duckweed in the production of bioethanol, as animal feed, or novel applications are among those that may be considered. One way to approach these questions is through a feasibility study, which places an idea into a context, allowing for careful deliberation and laying the groundwork for future implementation.

It is in this context that we worked to determine the feasibility of a duckweed-based phytoremediation system that contributes to the AUA Acopian Center for the Environment’s mission to reduce pollutant concentrations in Lake Sevan. We also studied the applications of duckweed harvests to ensure long-term opportunity development incentivizing continued collaboration with the local population. By completing these objectives, we will positively accelerate the efforts to remediate the Lake 16

Sevan ecosystem, as well as improve the quality of life in the communities in the surrounding basin. By lowering the concentration of pollutants feeding into the lake, the process of eutrophication occurring will be slowed and possibly stopped, preventing further damage from occurring to the basin’s ecosystem. This serves as the foundation for continuing efforts to raise the water level of the lake and impose more sustainable methods of industry in the surrounding area. The restoration of the Lake Sevan ecosystem also compounds the profitable nature of duckweed harvesting by revitalizing industry and improving economic development in the surrounding communities. 17

Background

In this chapter, we introduce Lake Sevan as well as the overarching problem of excess nutrient concentration in the lake that we seek to remedy. We then present the various pollutants affecting Lake

Sevan and give a brief overview of duckweed’s ability to decrease these pollutant levels by working as a phytoremediation system. The plant’s potential for opportunity development after harvest and the challenges associated with duckweed farming are also explored. Finally, we conclude by introducing the

AUA Acopian Center for the Environment, its goals, and the objectives of our research efforts.

History and Importance of Lake Sevan

Lake Sevan (Figure 1) is situated in Gegharkunik province, about 60 km (37 mi) north of Yerevan in Armenia’s volcanic highland (Babayan et al., 2006). Comprising one-sixth of Armenia’s territory, the lake is one of the largest freshwater high-altitude lakes in Eurasia and is considered one of the country’s most important natural resources, both culturally and economically. The lake itself has been a prominent

part of Armenian art, poetry, and music

for thousands of years, while the Sevan

Monastery, located near the shores of

Lake Sevan, is a lasting monument of

Armenian independence (Laplante, et

al., 2005). It is also Armenia’s main

source of irrigation water and is relied

upon as a source of fish, electricity,

tourism, and recreation. Twenty-eight

rivers and streams feed into Lake Sevan,

conversely only the Hrazdan River

drains Lake Sevan (Babayan et al., 18

2006). This outflow of water from the

River Hrazdan was regulated by the

Armenian Soviet Socialist Republic beginning in the early 1930’s to irrigate the Ararat Valley and control water levels for hydroelectric power generation (Lind & Taskakyan, 2005).

Naturally, the rate at which water left

Lake Sevan was 50 Mm³ (66 Myd³)

(Laplante et al., 2005). After some alterations were made by the Armenian

Soviet government, this outflow increased to approximately 700 Mm³ (915 Myd³) (Laplante et al., 2005). As the new outflow significantly outweighed the natural inflow, there was a drastic decrease in water level totaling 20 m, from approximately 1,916 m (6286 ft) above sea level in 1928 to 1,896 m (6220 ft) above sea level in 1998

(Figure 2). The surface area of the lake has also drastically decreased, diminishing 184 km2 , from 1,416 km2 (547 mi2) to 1,239 km2 (478 mi2) since 1933 (Lind & Taslakyan, 2005).

Due to this large reduction in water level, the soil previously covered by water has been exposed, leading to soil drying and salinization. This process has caused rising levels of erosion around the lake, leading to increased sedimentation rates. These factors, combined with other sources of pollution such as untreated sewage, agricultural runoff, and industrial strain, have destabilized the lake’s ecosystem enough to cause an accelerated form of eutrophication first observed in 1964. Promoting excessive organic growth in an aquatic ecosystem, eutrophication is a process that destabilizes the local food web through increased concentrations of nitrogen and phosphorus (Conley et al., 2009). Often caused by urbanization and agricultural expansion, this process causes high levels of oxygen deficiency via increased organic matter degradation (de Jonge et al., 2002). Oxygen deficiency in water can be problematic as all aquatic 19 organisms need dissolved oxygen to survive and a lack of oxygen can lead to dead zones incapable of supporting life. Algae blooms also block sunlight penetration into the water column and release toxins which harm the local aquatic life present in the body of water (de Jonge et al., 2002). These ecological changes endanger the 10 endemic species present in Lake Sevan (Table 1).

Table 1. Species Endemic to Lake Sevan (Babayan et al., 2006)

Type of Organism Scientific Name Common Name

Flora Acantholimon gabrieljanae Gabrielyan's prickly thrift

Flora Alyssum hajastanum N/A

Flora Astragalus shushaensis N/A

Flora Isotis arnoldiana N/A

Flora Isotis sevangensis N/A

Flora Ribes achurjani N/A

Fauna (Fish) Salmo ischchan Sevan trout

Fauna (Fish) Varicorhinus capoeta sevangi Sevan koghak

Fauna (Fish) Barbus goktschaikus Sevan barbel

Fauna (Bird) Larus armenicus Armenian gull

The over-exploitation of the lake’s natural resources has been a topic of national and international focus, and the Armenian government views the restoration of Lake Sevan as a high-priority issue. Much of the recent legislation to combat the deterioration of lake conditions is built upon the environmental regulation section of the Armenian constitution, which states Armenia “shall ensure the protection and reproduction of the environment and the rational utilization of natural resources.” A law referencing this aspect of the constitution, adopted in 2001 to “[establish] a legal and economic basis of state policy [for] natural development, recovery... [and] preserving… of Lake Sevan,”' has led efforts which have raised the water level to approximately 1900 m above sea level as of 2019. Other efforts (Appendix A), such as the

Lake Sevan Action Program (LSAP) developed by the Armenian Ministry of Nature Protection from 20

1996 to 1998, created a framework in which the water quality of Lake Sevan can be improved while still protecting the biodiversity present in the basin (Lind & Taslakyan, 2005). Although these efforts have created the structure for progress, little has been done to build from this legislation and successfully resolve the situation present in Lake Sevan.

Eutrophication, Metal Pollutants and Duckweed Water Treatment

Increasing phosphate and nitrate concentrations within Lake Sevan increase the likelihood of eutrophication, posing a major threat to the health of the lake’s ecosystem. These nutrient levels have reached concentrations well within eutrophication levels, and two large-scale cyanobacterial blooms in

2018 and 2019 have occurred as a consequence. Optimal concentrations (considered to be 0.457 mgN/L and 0.033 mgP/L respectively by Republic of Armenia Government Decision #75‐N) of these nutrients in the ecosystem are essential to steady growth. Outside influences such as fertilizer runoff and industrial waste add unnatural amounts of nutrients into the water, and manufacturing adds nitrogen to the atmosphere (Burt, 2019). In response to this increased concentration of nutrients, algae growth exceeds normal rates. This consequently has many negative effects on the surrounding ecosystem. Thick layers of floating algae on the lake’s surface completely block sunlight penetration, killing the aquatic plants below the layer’s surface. This, combined with the algae that eventually dies, creates large areas of dead organic matter. As this matter decomposes, bacteria and detritivores, which feed on dead organic material, consume significant amounts of oxygen. This process creates dead zones where other aquatic organisms cannot survive (Vinçon-Leite & Casenave, 2019). The results of these dead zones are not limited to ecological damage. The communities that depend on the body of water will incur an economic burden, and local fisheries will be impacted by the lack of diversity and reduced number of fish present in the lake's fish population. The increased algae growth on the lake’s surface results in a green-colored lake that decreases tourism revenue and lowers the cost of waterfront property. In many cases, the immediate solution to stop or reverse eutrophication is to reduce or halt nutrient flow into the water source.

Eventually the natural cycle of nutrients should slowly return to a steady state. However, if the cycle is 21 already compromised, then excess nutrients currently in the lake must be permanently removed in order to revert the process (McCrackin et al., 2017).

Another dangerous source of pollution is a high concentration of heavy metals. Excess concentrations of heavy metals in a body of water presents a great threat to the health of the surrounding ecosystem although water systems require a trace amount of certain metals used as nutrients. An influx of metals can occur when humans interfere with the natural state of the metal cycle (Martinez et al, 2020).

Examples of interference include fertilizer runoff, dumping of waste, and atmospheric pollution. These interferences are tied into the urbanization or development of a community. Urbanization can be considered the main cause of most metallic water pollution and the worsening of water quality. Due to the danger to water systems that urbanization carries, water quality is important to monitor. Water with a high concentration of metals has the capacity to damage both the ecosystem near the water and the human community that relies on the system (Pal & Maiti, 2019). Once metals enter the system, plants intake them through the water, regardless of the toxicity of the metal. Cadmium is one example of this, as it doesn’t play a role in plant growth and is toxic at high levels (Martinez et al., 2020). The effects of high metal concentrations are not limited to the local ecosystem. Local communities are also harmed by feed from the lake that their livestock consumes have high metallic content. Similarly, any fish harvested from the water source will also contain some trace metals that humans will eat directly, potentially causing a number of neurological and cardiac problems (Pal & Maiti, 2019). One method of permanently removing both heavy metals and excess nutrients from water systems is to harvest plants that are growing and taking up the pollutants. One such example of a plant that can be used for this is duckweed.

Duckweed is a family of floating freshwater plants that consists of over 40 species. All of the species share similar characteristics and growing conditions. As pictured in Figure 3, duckweed grows in clusters of leaves called fronds. These fronds grow off of each other and number between two and six for each duckweed plant. The leaves on duckweeds are round and about one millimeter to a centimeter

(1/25th to 2/5th of an inch) across with faint veins running across the surface (Blocks & Rhoades, 2011).

A small root system connects the leaves and stabilizes the plant in shallow water. The length of the root 22 varies by species and nutrient level (Leng, 1999). Duckweed does have flowers despite reproducing asexually (Blocks & Rhoades, 2011).

Duckweed is a floating macrophyte native to temperate, still water sources. The water surface temperature range is 6°C to 33°C (43°F to 91°F); however, the optimal temperature for duckweed is 27°C

(81°F) (Blocks & Rhoades, 2011; Papadopoulos et al., 2014). Ideal duckweed growing conditions are not necessary for duckweed to grow; however, duckweed grows considerably slower when resources and conditions are poor. Duckweed can tolerate approximately 2.5% salt level in water, but lower levels are ideal. The optimal pH for duckweed ranges between 6.5 and 7 but it can survive in pH levels between 5 and 9 (Leng, 1999). Duckweed can grow in water depths as low as muddy water or out on the surface of deep water. Due to the labor that duckweed farming requires, growing duckweed is typically done in water less than half a meter (1.6 ft) deep (Blocks & Rhoades, 2011). 23

Floating macrophytes are any plant that grows in water without the need to root into soil. Some common examples are water lilies and watermeal. The unique characteristics of floating macrophytes make them a viable option for a floating macrophyte farm. A floating macrophyte farm is when floating plants absorb the excess nutrients and are harvested to remove them from the cycle (Körner et al., 2003).

The most important aspect to floating plants acting as nutrient and pollution sinks is the surface area that these plants take up. Duckweed fronds create mats (Figure 4) that sit on the surface of ponds, which provide a large amount of surface area that allow the biofilm to collect nutrients and dissolved sediment from the water. These nutrients in a wastewater treatment system are the excess nitrogen compounds or the metals that are polluting the water source. The root system in floating plants provides more advantages

(Tanner & Headley, 2011). Floating plants do not use roots to provide a foundation. The sole purpose for the roots is to increase surface area in the water. As a result, floating plants can survive in still water no matter the depth, unlike marsh plants. This unique feature also allows for low effort relocation of the plants, whether for installation or for harvest (Tanner &

Headley, 2011). Under the previously described conditions, duckweed can grow rapidly. By harvesting the duckweed at an appropriate rate, the nutrients that are absorbed by the duckweed can be permanently removed from the water system. By taking these nutrients out of the system, it is possible to slow and reverse the process of eutrophication (Papadopoulos et al., 2014).

Besides its effectiveness as a wastewater treatment system, duckweed and its applications present the opportunity for economic development. Due to its protein content, which ranges from 15% to 45% dry weight, duckweed has been used as a source of sustainable livestock feed (Cheng & Stomp, 2009). With a 24 fast growth rate and an ability to thrive in a wide variety of environments, duckweed is able to supply a large proportion of needed protein required for animals with no adverse effects (Cheng & Stomp, 2009).

Duckweed farms also have a limited footprint with regards to CO2 emissions. To obtain 1 kg (2.2 lbs) of duckweed, only 0.4 kg (0.9 lbs) of CO2 is produced (Liu et al., 2021). In order to increase the protein content cultivated, water with high levels of nitrogen concentration should be considered in order to stimulate amino acid synthesis (Liu et al., 2021).

Duckweed can also be harnessed for its starch content, which ranges from 3% to 75% dry weight depending on various growing parameters (Cheng & Stomp, 2009). Higher starch content presents the possibility of using duckweed for industrial feedstock in bioenergy production (Liu et al., 2021). Unlike fossil fuels, the consumption of ethanol-based biofuel is considered atmospheric carbon neutral due to the consumption of CO2 that occurs during biomass growth. It is also able to biodegrade without environmental damage, reducing the hazard of spills. Ethanol is an alcohol that is most commonly developed through the fermentation of biomass feedstocks with high starch and sugar contents. Cellulosic ethanol can also be produced through more complex production processes, but these are beyond the scope of this project due to their high production costs (Thompson, 2010). Besides being used as a fuel and an additive to gasoline, ethanol has other applications in medicine, cosmetics, and household products.

Challenges with Maintaining a Duckweed Farm

Duckweed, being one of the fastest growing plants in the world, is ideal for use in phytoremediation and wastewater treatment; however, this characteristic presents a unique set of challenges when maintaining duckweed farms (Leng, 1999). With the ability to double its biomass every

16 to 48 hours under optimal growing conditions, duckweed grows significantly faster than other aquatic plants. Plants such as mosquito fern or water fern double their biomass every five to six days in comparison (Miranda et al., 2016). Due to the rapid rate of growth, duckweed harvests need to be performed every few days to ensure that farms are kept under control. If duckweed is allowed to grow without routine maintenance, it will form thick mats that prevent sunlight penetration into the surrounding 25 water column. This lack of sunlight causes dead zones below the duckweed mats where other plant species cannot conduct photosynthesis and sustain life. The large-scale death of aquatic life in the water reduces the dissolved oxygen levels in the water, as there is less oxygen production. Compounding this is the decomposition of dead biomaterial that occurs, in which bacteria levels increase and consume the remaining dissolved oxygen. This cyanobacterial boom also increases the release of CO2 in the water

(Gupta & Prakash, 2014). With a sudden decrease in oxygen, a eutrophication-like event occurs, creating a dead zone in the vicinity. The risk of this occurring is highest in small, enclosed bodies of water such as ponds as the water is more stagnant, reducing the ability for organisms to travel to healthier growing conditions. Smaller changes in nutrient levels affect a greater portion of the pond than it would in a larger body of water such as a lake. While the risk of algae being killed by overshadowing duckweed mats is minimal in a river system, where the water is constantly moving, this must still be considered to prevent the duckweed from overshadowing and killing itself. Frequent harvesting will also lower the risk of excess duckweed being washed downstream into Lake Sevan and growing in unwanted areas.

To control and counteract unwanted growth, there are four effective methods of reducing and removing duckweed from unwanted growing areas: manual removal of the duckweed, limiting excess nutrients in the water, overshadowing, and dredging. Manual removal of duckweed is a labor-intensive process that must be performed thoroughly, as missing a few duckweed plants will allow for the duckweed to quickly repopulate the area. Limiting excess nutrients in the water drastically reduces the reproduction rate of duckweed as duckweed thrives in nutrient-dense waters (Leng, 1999). In order to do this, sources of nutrient input into the water system from farms or factories must be identified and limited.

Unfortunately, within the scope of this project, this is an impractical approach of duckweed control as it would be counterintuitive. If the main sources of nutrient input in Lake Sevan were reduced to eliminate unwanted duckweed growth, there would be no need to grow duckweed to reduce excess nutrients in the water if the nutrient input can simply be reduced.

Dredging and overshadowing are two other ways to control and remove unwanted duckweed growth by directly altering the growing conditions (van den Berg et al, 2015). Dredging involves the 26 removal of the bottom layers of the waterbody to increase the water depth. To use it to limit duckweed growth, the river or lake would have to be dredged to a depth ill-suited for duckweed growth. Despite dredging being effective at removing duckweed it is not an advisable method as the ecosystem within the dredged area is destroyed, and any pollutants locked in the sediment are released into the water. On top of this, it is an expensive process, with costs ranging between $4 to $8 per cubic yard (about $5.24 to $10.46 per cubic meter), as well as setup and shutdown costs ranging from between $20,000 to $50,000

(Dredging 101, n.d). Overshadowing is a method of duckweed control that involves creating shade over the areas of water that the duckweed is growing in, cutting off its access to sunlight. This can be achieved by placing platforms directly in the water on top of the growing duckweed or trees can be planted along the riverbank to provide shade over the water. While the lack of sunlight will eliminate the duckweed, it will also impact the growth of any other photosynthetic organisms in the area. This risks causing an eutrophication-like event and the creation of a dead zone similar to that caused by thick duckweed mats.

While the shade can be removed to help prevent a eutrophication-like event from happening, the removal of the shade allows for the potential of duckweed to regrow, requiring further efforts to control its growth

(van den Berg et al, 2015).

The best practice to avoid eutrophication-like events and the creation of dead zones caused by uncontrolled duckweed growth, as well as lowering the likelihood of duckweed spreading to unwanted areas is to harvest duckweed often. Harvest requirements depend on the growth rate of the duckweed, the maximum size of the farming area, and the mat density left unharvested. Harvesting can be done manually using tools such as rakes or using automated systems that suck up surface water and duckweed, depositing duckweed in a catch bin and returning the water back to the pond.

Regardless of the method of harvesting, freshly harvested duckweed consists of 92-94% water and must be dried before being used in most applications. There are two common methods of drying duckweed, solar drying and oven drying (van den Berg et al, 2015). Solar drying is a cheap drying method as the duckweed simply needs to be harvested and placed onto black plastic sheets and allowed to dry in the sun for several days (Leng, 1999). The costs of solar drying are minimal, only requiring black plastic 27 sheets and drying can take place on site. The alternative method of drying involves placing the duckweed into ovens, shortening the drying process from a few days down to a few hours. If oven drying is being used, then it must take place off site adding transportation costs on top of the costs of running ovens for hours on end.

As mentioned previously, metal and metalloid contamination in water presents a risk of indirect damage to the community which relies on this water. Any plants or animals grown in the contaminated water or fed with contaminated feed will also become contaminated in the process of biomagnification.

This will extend to any humans who consume contaminated duckweed directly or animals fed with contaminated duckweed. These people potentially face serious health risks of acute or chronic metal toxicity, with long-term exposure potentially leading to neurological and/or physical degeneration (Engwa et al., 2019). The exact health effects will vary metal to metal and on the concentration of the metal.

For example, at low levels arsenic poisoning can lead to nausea, vomiting, and reduced erythrocyte (red blood cells) and leukocyte (immune cells) production. Long-term and acute exposure to this metalloid can lead to the destruction of blood vessels and gastrointestinal (GI) tissue as well as various cardiac and neurological problems. Iron poisoning mainly affects the GI tract causing vomiting,

GI tract bleeding, GI ulcers, as well as cardiac problems such as tachycardia and hypotension (Engwa et al., 2019). Both arsenic and iron are commonly found pollutants in Lake Sevan. Arsenic has been found at concentrations as high as 9 or 10 μg/L in several locations, while iron concentrations have been found as high as 460 μg/L, which exceeds the recommended level for drinking water (Aghajanyan et al., 2018). As metals present a great health risk, whenever duckweed is harvested, the metal contamination level must be determined so that the duckweed can be used in an appropriate way.

About the AUA Acopian Center for the Environment

The AUA Acopian Center for the Environment is a branch of the AUA whose goal is to “promote the protection and restoration of the natural environment through research, education, and community outreach.” In an effort to further their goal, the Acopian Center currently manages twenty-one total 28 research and community outreach programs that focus on a wide variety of areas. These include environmental policy, sustainable natural resource management, information technology, and the interaction between the natural environment and the man-made surroundings. The duckweed phytoremediation project proposed and sponsored by the Acopian Center is tied directly to the Center’s focus on sustainable resources management and supports other ongoing projects examining the impact sulfur cycling has on eutrophication and waste quantity and governance management. The restoration and continual upkeep of Lake Sevan is of utmost importance to the AUA Acopian Center as well as the

Armenian Government, and this project serves to assist in the preservation of Lake Sevan (Acopian

Center for the Environment, n.d).

Summary

With the risk pollution poses to the delicate ecosystem of the Lake Sevan Basin and the communities that rely on its resources, the AUA Acopian Center for the Environment is seeking to implement community managed duckweed farms in several key locations around the Sevan Basin. This system will help prevent phosphate and nitrate concentrations in Lake Sevan from continuing to rise as well as reducing the inflow of other pollutants into the lake such as arsenic and iron. Allowing for further methods to be implemented, the system counteracts the eutrophication process occurring in Lake Sevan.

The duckweed-based phytoremediation project also presents opportunity for the surrounding populace, as duckweed harvests may be profitable through ethanol or animal feed production. The restoration and continual maintenance of the lake’s natural resources also allows surrounding industries to thrive as the resources they depend upon will no longer be threatened by pollution related complications. We are working with the AUA Acopian Center for the Environment to determine the feasibility of a duckweed-based phytoremediation system that will contribute to the Center’s efforts to reduce pollutant levels in Lake Sevan and economically benefit the surrounding populace. 29

Methodology

During this project we determined the feasibility of a duckweed-based water treatment system that contributes to the AUA Acopian Center for the Environment’s mission to reduce pollutant concentrations in Lake Sevan. We also conducted an analysis of the potential economic opportunities created from duckweed harvests in order to incentivize long-term involvement and to help ensure the project remains financially viable. As we worked towards completing both goals, we focused on accomplishing the five following objectives:

1. Analyze geographic and water quality data from Lake Sevan’s tributaries to identify three

locations most suitable for and in need of duckweed-based phytoremediation.

2. Identify logistical challenges that must be addressed and determine local attitudes

towards pollution in Lake Sevan and the implementation of a duckweed-based

phytoremediation system.

3. Perform an opportunity analysis of duckweed farming in the Lake Sevan region,

including an economic assessment of the costs of establishing and maintaining duckweed

farms and the potential profit duckweed harvests can provide for its farmers.

4. Design a small-scale pilot farm to help the Acopian Center determine the effectiveness

and feasibility of implementing duckweed farms in Lake Sevan.

Each objective contributed to our general understanding of the situation and ability to design the pilot. Once we gathered enough information to make an informed recommendation for our sponsor, we began to design a small-scale pilot project as outlined in Objective 4. The completion of each of these objectives aided us in determining if duckweed is a viable phytoremediation system in Lake Sevan, as well as benefiting our sponsor in their overall endeavor to reduce pollution and maintain the Lake Sevan ecosystem. 30

Objective 1. Geographic and Pollution Data Analysis

In order to maximize the effectiveness of duckweed as a phytoremediation system, several of the rivers feeding into Lake Sevan were assessed to identify locations that could maintain duckweed growth and reduce nutrient flow into Lake Sevan. For each tributary researched, emphasis was placed on the mouths of the rivers as they are often the most polluted sections of the river, and the pollutants found at the mouths of the river directly enters Lake Sevan. Using water quality data collected for a report conducted under the European Union Water Initiative Plus for the Eastern Partnership countries

(EUWI+), ten tributaries of Lake Sevan were assessed on nitrate and phosphate concentration, overall nutrient concentration, and the National Sanitation Foundation’s Water Quality Index (WQI) standards.

The National Sanitation Foundation (NSF) is a global, independent organization that develops public health standards in order to protect and improve human health. Their WQI is one of the numerous water quality indices that have been developed as a convenient means of summarizing water quality data (Wills et al., 1996). In order to conduct the analysis, yearly averages of each water quality parameter from 2010 to 2019 was calculated using the EUWI+ data. Once this step was completed, each assessment was conducted by ranking each parameter on a scale of one to ten, one being the highest level or worst condition, and ten denoting the lowest levels of pollution and best water quality. The rankings were based upon a year-by-year comparison with the values from each of the other ten rivers. Once this process was complete, an average total ranking of the rivers over the ten year period was determined for each assessment.

The initial analysis of the provided nutrient content data ranked the tributaries in terms of their average yearly phosphate and nitrate concentrations. Due to the importance of reducing nitrate and phosphate inflow into Lake Sevan in order to prevent eutrophication, as well as their positive effect on duckweed growth at higher concentrations, this ranking was weighted heavily in the final determination of viable farm locations by a factor of 0.5. Rankings for each nutrient and year were calculated, and from this an average of the rankings for each river was determined for both of the nutrients. The final ranking 31 for this step averaged each river's ranking for phosphate and nitrate concentrations, creating a single value for each river.

The output from the next assessment determined the average concentrations of various nutrients important for duckweed growth. These nutrients include phosphorus, nitrogen, iron, magnesium, calcium, potassium, and sulfur. Each average yearly nutrient concentration was ranked for each river in order to calculate an average final ranking for each nutrient. These values were then used to compute a final holistic ranking for nutrient concentrations vital for efficient duckweed growth. Due to the importance of having a high duckweed biomass yield during harvest for use in other applications, this ranking held a weight of 0.35 in the final ranking calculation.

The final step before determining the final ranking of locations in which we would recommend implementing a duckweed based phytoremediation system was to rank the tributaries based on their NSF

WQI scores. Scores were calculated based upon the total average parameter calculations made from the yearly averages computed previously. This outputs one WQI score for each of the rivers, which were compared and ranked accordingly. Parameters of interest were the river’s dissolved oxygen, pH,

Biochemical Oxygen Demand (BOD₅), total nitrate concentration, total phosphate concentration, and total solids. WQI scores are calculated by determining a Q-value from each parameter and weighing this parameter accordingly (Table 2).

Table 2. NSF Water Quality Index Parameter Weights Factor Weight

Dissolved Oxygen 0.17 pH 0.11

Biochemical Oxygen Demand 0.11

Total Phosphate 0.10

Total Nitrate 0.10

Total Solids 0.07 32

Q-values are required as they convert the values found from each individual parameter into a standardized value. This helps to interpret water quality data, as the higher the Q-value the better the quality of that specific parameter is. Due to there being fewer than the nine measurements usually used to calculate the WQI value, the final value was scaled accordingly to maintain a range from 0 to 100. Once the WQI values were calculated, scores were compared and rivers were given ranks accordingly. When calculating the final rankings for recommended duckweed farm locations, these rankings were weighed with a value of 0.15, the lowest of all three initial rankings. This is due to the fact that some factors, including BOD₅ and dissolved oxygen, are important to the health of the river’s ecosystem but not that of duckweed growth. With the initial rankings completed, a final ranking was determined based upon the weights previously established. The initial nitrate and phosphate concentration ranking was multiplied by

0.5, while the initial duckweed growth nutrient concentration data and WQI score rankings were multiplied by 0.35 and 0.15 accordingly. These values were then summed for each river and ranked from lowest to highest.

Once the rivers had been ranked, each was looked up using Google Maps and Google Earth to find towns located near each river. From this list, the population of various municipalities was found in the regions surrounding each of the three rivers selected. These communities of varying sizes were looked at on Google Maps again to find which ones are located near to the mouths of the rivers. Eighteen towns were selected for in-person interviews to be conducted, and from this list a final selection was made which took into account travel considerations. In addition to using Google Maps to determine town locations, a topographic map of Armenia was found and used to find the approximate elevations of the area surrounding each of the towns. This was done to see if the topography of the region would potentially impact duckweed farming. This consideration must be made in order to identify logistical issues that may arise during implementation and maintenance.

Finally, using hydrology data provided to us by Alexander Arakelyan, a hydrology expert and project coordinator at the Acopian Center, calculations were made in order to determine the total mass of nitrogen and phosphorus flowing into Lake Sevan from the rivers identified for implementation. The 33 initial calculation involved determining the yearly average concentration of phosphorus and nitrogen in the year 2019 for each river. From this, a total yearly volume flowing out of each river was determined using average volumetric flow rates. Using these two calculations, a yearly total of phosphorus and nitrogen flowing out of each river could be calculated. Due to limitations in concentration data, an estimation for nutrient concentration data for the Martuni River had to be conducted. 34

Objective 2. Assessment of Local Knowledge and Attitudes

Beyond the initial development of the duckweed farms, the AUA Acopian Center for the

Environment will not be involved in maintaining the farms or performing the daily work required. The responsibility for this work will fall upon local farmers who will maintain the farms and ensure the sustainability of the process. Because the local populace will be primarily responsible for the duckweed farms, it was important to assess their level of awareness of environmental issues and their overall attitude towards duckweed farming. Interviews based on several guiding questions (Appendix B), were conducted with local municipality leaders to gain a better understanding of their communities’ economic structure, their feelings toward farm implementation, and their concerns and suggestions for further consideration.

These individuals were considered representative of their community’s general populace and the interviews were meant to be informal, as a transcript was not kept of the conversation. Local business owners in industries relevant to duckweed applications, including fishery managers and farmers, were also interviewed in order to gage their interest in the economic opportunities created from duckweed farming. By doing so, our sponsor gained insight into which applications of duckweed biomass will incentivize the most collaboration within these municipalities. As travel restrictions prevent us from being able to physically conduct interviews ourselves, Hovhannes Torosyan, a student at the AUA, assisted the team in conducting interviews with municipalities representatives. Questions were asked of the representatives regarding their current sources of animal feed, the requirements they need met to consider a duckweed-based feed as a feasible replacement, as well as the infrastructure present in their establishment. Further efforts were made to determine if these individuals would be interested in committing resources towards farm implementation. By gathering this information, we were able to determine the general consensus of the municipality towards water quality issues and their interest in the project. Further decisions regarding community involvement and implementation locations can be supported based upon this information. 35

The reports created from these interviews outlined the general feelings toward the project, highlighting specific concerns within the community and logistical and legal considerations that would need to be addressed if further efforts are made towards implementation in the area. Information was broken up into four sections: Pros, Cons, Important Considerations, and Miscellaneous. The “Pros” section was used to record information about the specific municipality that supported our project efforts, such as ideal terrain or general interest in duckweed applications. The “Cons'' section was for information that would hinder farm implementation in the area, such as a lack of resources or a negative view of project goals. The third section, titled “Important Considerations”, was used to hold information and issues raised that would need further consideration, altering our approach towards implementation in the area. An example of this would be the risk of flooding in the area or current projects being conducted in the community. Finally, “Miscellaneous” was used to hold other information unique to the municipality, such as the prevalence of communintity members migrating for work, or recommendations from the interviewee for further consideration. Through these reports, a framework in which further communication can be built upon was developed for each of the surrounding communities. 36

Objective 3. Opportunity Analysis of Duckweed Farming

A key aspect of our project was developing an understanding of the opportunity created by duckweed farming in the surrounding communities of the Lake Sevan Basin. In order to achieve the desired ecological impact of our project, we explored profitable ventures within duckweed farming for the local populace that may assist in incentivizing participation in the development and maintenance of duckweed farms by locals. Without continued support of the local populations, the project is at risk of being abandoned. Without routine harvesting and general upkeep of the farms, duckweed could further threaten the balance of Lake Sevan’s ecosystem and grow without bounds. Also, without outlining possible usages of duckweed after harvest, a detrimental situation will arise where the duckweed begins to decay and further increases damaging nutrient content in the lake water. As part of the effort to incentivize duckweed farming, we aimed to assess various options of post-harvest products of duckweed by evaluating the risks and opportunities for a duckweed product’s development when measured against the goals of our project stakeholders. Part of this analysis was done using an algorithm to estimate the amount of biomass that can be produced growing Lemna gibba or Lemna minor in the three rivers identified in Objective 1. Biomass estimation reveals the possible nutrient uptake, the feasibility and impact of various applications, as well as the growth period and harvest timeline required.

The biomass algorithm was based upon a growth rate model determined by equations from Lasfar et al. (2007), McLay (1976), Frederic et al. (2006), and Peeters (2013). The main equation developed for the model was used to calculate the intrinsic growth rate, which would include functions for water temperature, phosphate and nitrate concentrations, pH, and photoperiod in the various locations. Growth factors for each of the variables were determined in order to develop the final intrinsic growth rate equation. For phosphate and nitrate concentrations, the growth factor was dependent on nitrate and phosphate saturation and inhibition growth constants, as well as the nutrient concentrations present in each river. The saturation growth constant accounts for the effect of saturation that occurs at low nutrient concentration levels, while the inhibition growth constant accounts for the effect of inhibition that occurs 37 at higher nutrient concentrations. These values were calculated via mathematical regression (Lasfar et al.,

2007). The growth factor for temperature is based upon two non-dimensional constants previously determined by Lasfar et al. (2007) and the optimal and recorded water temperature. Due to the lack of water temperature data for each of the three rivers, data were gathered from Big Sevan, the larger of the two subsections of Lake Sevan, and considered a relative estimation to the water temperature present in the tributaries at the same time. For L. minor and L. gibba, the optimal temperature was considered to be

26℃ (79°F). The third variable in which a growth factor was determined was photoperiod, the period in which the duckweed would receive sunlight. Rather than calculating photoperiod as suggested by van den

Top (2014), previously recorded data regarding the average monthly hours of sunlight in the Sevan region for each month were utilized. This growth factor was based upon previously calculated non-dimensional constants from Lasfar et al. (2007) and the recorded and optimal photoperiods. For the two duckweed species of interest, the optimal photoperiod was considered to be 13 hours. Finally, the growth factor for pH was determined using a 5th degree polynomial equation modeled by McLay (1976). A variable representing the growth dependence on pH was calculated, and the intrinsic growth rate depending on pH utilized this value as well as the constants for maximum intrinsic growth rate and maximum intrinsic growth rate depending on pH.

From these initial growth factor calculations, a global model was developed combining the different growth factors in order to calculate the global intrinsic growth rate. This equation incorporates a maximum intrinsic growth rate constant as well, which is based on maximum growth factors, in order to limit modelled growth to a maximum. From this intrinsic rate, the specific growth rates and time-dependent mat density could be calculated. For specific growth rate, an equation developed by

Frederic et al. (2006) was utilized depending on the intrinsic growth rate as well as the limit mat density, initial mat density, and time between harvest periods. The time-dependent mat density was determined by

Lasfar et al. (2007) and involved integrating the density of the duckweed area over time. From this, an equation depending on mat density limit, initial mat density, intrinsic growth rate, and harvest period could be utilized. 38

This model allowed an estimation of total dry biomass harvested from March to October for the

Gavaraget, Masrik, and Martuni rivers. This time frame was selected as the minimum temperature limit to which L. gibba and L. minor can grow is 4℃ (39°F), which was greater than the average water temperature of Big Sevan in the months outside of this range. Using a mat density limit of 176 kg/m² and an initial mat density of 100 kg/m², which were determined by research conducted by Rooijakkers (2016), values for pH and nutrient content for each month were inputted into the model for each river. Each growth factor was calculated and a global intrinsic growth rate was determined. From this, the harvested biomass was found by multiplying the number of harvests, determined by the optimal period of time between harvests and the number of days within the month, and the time-dependent density minus the initial mat density. Optimal harvest times were estimated based on which period led to the greatest biomass output. It was determined that March and April would only have harvests at the end of the month

(31 and 30 days respectively), while the period would be reduced to three days between harvests for the months of May through October. Harvests for each of these months would also occur on the final day of each month, to ensure that the initial mat density was 100 kg/m² (20.5 lb/ft²) for every month and growth rates were optimized. After estimates were made for each month over the time span, a total biomass estimation was compiled by adding together each month’s biomass output. This was done for each of the three rivers.

Much of the information gathered in order to perform the opportunity analysis gathered through an extensive literature review. Our focus was on previous research that looked into duckweed being used as an input in animal feed, fertilizer, and bioethanol production. When looking at previous research into the use of duckweed as animal feed, research was conducted in three main stages. The first stage was a general overview of the literature looking at what animals can potentially be fed duckweed. The second stage of research looked further into using duckweed as fish and poultry feed, as initial research showed fish and poultry were the two most common animals to be fed with duckweed. During the second stage, focus was primarily placed on identifying the advantages of feeding fish and poultry with duckweed in comparison to commercial feeds or when used in tandem with commercial feeds as a supplement. After 39 finding that duckweed as animal feed offers advantages over commercial feed in terms of promoting growth, the third stage of research began by looking into negative effects associated with using duckweed as animal feed as well as any processing that is needed before duckweed is in a suitable form to be given to animals. A similar three-stage research approach was used when assessing the viability of using harvested duckweed as a fertilizer or as an additive for commercial fertilizer. Initial research was done looking at the benefits of using duckweed as fertilizer and a fertilizer additive in terms of improving soil quality, improving crop yields, and improving crop quality. The second focus of research was placed on how using duckweed can negatively impact the growth of crops. The final focus of research regarded what processing methods duckweed needs to undergo before it is used as fertilizer and potential storage options for the duckweed so that duckweed can be stored for future use either as animal feed or fertilizer.

In addition to researching the applications and benefits of using duckweed as feed and fertilizer, research was also conducted into ways to measure potential heavy metal contamination in duckweed to avoid poisoning livestock and crops.

Bioethanol research was broken into three parts: methods of starch accumulation, biomass output, and finally bioethanol production and output. Information regarding optimal conditions for starch content development and the effects of various growth parameters on starch development were gathered, mainly from previous pilot farms developed for the purpose. Methods to induce starch development post-harvest were also found, including alterations to nutrient content to induce nutrient starvation and salinity changes in the growing environment after initial harvest. During research focused on biomass output, we considered the previously found methods of starch content development and looked to determine their effectiveness. Data from the previous pilot studies highlighted the effectiveness of nutrient starvation and the introduction of NaCl, while displaying the percentage of starch content feasible under certain growing conditions. These data were then compared to the growing conditions in the three tributaries to estimate the starch content that could be expected for the various species implemented using our biomass model. Once an estimation of both the total dry biomass output and starch content present was developed, focus was shifted towards the ethanol production process and the requirements needed in 40 order to make it a profitable venture. The main procedure that was considered was the formation of a hydrolysate through enzymatic hydrolysis and then fermentation of said hydrolysate in continuously stirred tank reactors (CTSR). From previous research, information about the production procedure and required infrastructure was found. These studies also determined the ethanol yield possible based upon starch content, as well as comparing this yield to that of conventional bioethanol feeds such as maize. All of this information was used to estimate the ethanol yield feasible under the conditions found within Lake

Sevan’s tributaries to determine if this application is a profitable venture worth investing in. 41

Objective 4. Design of a Duckweed Pilot Farm

The design of a pilot duckweed farm, later to be implemented by the AUA Acopian Center for the

Environment, was the final objective we focused on. This endeavor relied on much of our previously completed research to guide us in our design. While completion of our objectives helped us to identify potential locations for the farm and logistical considerations, our opportunity analysis undergirded the design of the pilot farm. Because the feasibility of duckweed applications remains unknown, it is essential that the pilot farm helps generate more accurate data regarding duckweed growth in Lake Sevan. The generation and gathering of these new data will help the AUA Acopian Center clarify the amount of biomass production possible, the rate at which duckweed uptakes nutrients, and how to refine the initial opportunity assessment based on the actual biomass production and composition.

As the actual building of the pilot farm and the testing performed on the farm is beyond the scope of our project, we designed the pilot farm to be eventually developed in the Lake Sevan region. The design of the pilot farm was informed by our research and includes the following:

1. Determining what species of duckweed(s) should be grown on the pilot farm

2. Selecting a suitable location for the farm.

3. Determining an appropriate size for the farm.

4. Considering potential layouts including any infrastructure that may be needed.

5. Finding methods to measure nutrient and heavy metal uptake by the duckweed.

6. Recommending cost-effective harvesting and drying techniques.

To begin the design process, a literature analysis of the growing conditions of various duckweed species was conducted. Primary analysis focused on 15 duckweed species referenced in Elias Landolt’s

The family of Lemnaceae- a monographic study Volume I. This paper references several studies performed in the mid-to-late 1900’s by various researchers including Landolt himself. While this paper includes background information on duckweed, including morphology and karyology, which is the study of the nucleus (specifically chromosomes) of duckweed, the focus of our research was placed on the habitat 42 demands of duckweed. Of particular interest was a study Landolt conducted in Zürich during the winter of

1957-58. After a cold winter in which the basins the studied duckweed grew in froze over, Landolt noted seven species that were able to regrow in the spring. After identifying the species that were able to do so, further research about minimum growth temperatures for the seven duckweed species was done to select the three species with the lowest minimum growth temperature ranges (Landolt, 1986). When it was found that L. minor, L. gibba, and Lemna trisulca had the lowest growth temperature, research into the average starch and protein content for each species was done to help decide what post-harvest products were most viable. It was also conducted to see if any of the three duckweed species offer distinct advantages or disadvantages for use in various post-harvest products.

In addition to researching the most promising duckweed species to use in the pilot farms, further research was done to see what effect monocultures (the cultivation of a single species in an area) and polycultures (the cultivation of multiple species in an area) has on the growth of duckweed. Literature could not be found specifically using the three species selected as polycultures in duckweed farms.

Rather, there was literature found regarding the general growth of duckweed as both polycultures and monocultures, offering various advantages and disadvantages. While polycultures, when compared to monocultures, seemed to have little impact on the actual growth rate of duckweed, polycultures did seem to impact the protein levels of duckweed. Further research into this revealed that polycultures do impact protein levels along with starch levels and the nutrient absorption rates of duckweed. Using the results of the polyculture research, along with the information gathered from researching the properties of various duckweed species, we could recommend a combination of duckweed species to maximize the timeframe of duckweed growth and provide the highest quality duckweed for various post-harvest applications.

Other research was done looking at other duckweed farms used in conjunction with pig farms to reduce the nutrient contents in wastewater produced by the swine farm, as well as looking at ways to divert the flow of the rivers. Various water diversion methods were looked at, mainly involving the development of pits near the river and creating canals to connect them. The building of dams to slow the river and divert the flow was also researched. After conducting the literature review, the data was 43 collected and compiled along with the other data collected from working on the other three objectives.

After this, the final iteration of our pilot farm design was planned out, starting with the broadest features, such as where the pilot farm should be built. We then moved into specific design recommendations, including what species of duckweed will be grown and ways to determine the heavy metal concentration of harvested duckweed. The information of the final design was presented as a blueprint created using

SolidWorks. The blueprint presented is the cheapest and simplest design for the potential farm. We have recommended systems that could be used to upgrade the farm, dependent on increased funding for the farm. Accompanying the model is a written description of the suggested duckweed species to use in the pilot farm. We have also included recommended methods of duckweed harvesting and drying, with further recommendations on how to upgrade both processes. 44

Results

The desired result of this project was to assist the AUA Acopian Center for the Environment in determining the feasibility of a duckweed phytoremediation system implemented in the most polluted tributaries of the Lake Sevan Basin to prevent further eutrophic events in Lake Sevan (Figure 5). As part of this process, hydrology and water quality data was analyzed in order to determine locations most suitable for project implementation. Once locations were chosen, local attitudes from the surrounding municipalities were determined to gauge project interest and reveal concerns and logistical challenges.

After, the examination of the potential applications of post-harvest duckweed biomass and estimation of the biomass yield was also completed to give an overview of the economic benefits possible after implementation of the duckweed phytoremediation system. Finally, an initial design for a duckweed pilot farm was developed to allow the AUA Acopian Center to possibly implement the pilot. With these deliverables representing our end goal, we outlined objectives that would allow us to use a variety of methods to bring about the results that we desired. 45

Objective 1. Geographic and Pollution Data Analysis

The work we conducted focusing on our first objective used nutrient concentration and water quality data previously collected for a report conducted under the European Union Water Initiative Plus

(EUWI+). Rivers were ranked based upon their phosphate and nitrate concentrations, their ability to support optimal duckweed growth, and their NSF WQI score. The Gavaraget, Martuni, and Masrik rivers were found to have the highest average nitrate and phosphate concentrations of the ten rivers, a ranking which held a weight of 0.5 when computing the final rankings (Table 2). When it came to rivers which presented conditions optimal for duckweed growth, a ranking with a weight of 0.35 in the final rankings, the Gavaraget, Masrik, and Sotk were the rivers that were determined to optimize duckweed growth.

These rivers had pH levels similar to optimal standards as well as high nutrient concentrations, including nitrogen, phosphorus, sodium, and others analysed, which support duckweed biomass production. Finally, the Gavaraget, Sotk, and Masrik rivers were found to have the lowest NSF WQI scores based upon a variety of water quality factors (Table 2). This ranking received a weight of 0.15 in the determination of the final rankings.

Table 3. Water Quality and Nutrient Concentration Rankings Phosphate and Nitrate Concentration Ranking (Weight: 0.5)

River Gavaraget Martuni Masrik Shoghvak Tsakkar Vardenis Argichi Sotk Karchaghbyur Dzknaget

Ranking 1 2 3 4 5 6 7 8 9 10

Optimal Conditions for Duckweed Growth (Weight: 0.35)

River Gavaraget Masrik Sotk Martuni Shoghvak Tsakkar Argichi Vardenis Dzknaget Karchaghbyur

Ranking 1 2 3 4 5 6 7 8 9 10

Water Quality Index Ranking (Weight: 0.15)

River Gavaraget Sotk Masrik Shoghvak Tsakkar Martuni Vardenis Argichi Dzknaget Karchaghbyur

Ranking 1 2 3 3 3 6 7 8 9 9

Overall Ranking

River Gavaraget Masrik Martuni Shoghvak Tsakkar Sotk Vardenis Argichi Karchaghbyur Dzknaget

Ranking 1 2 3 4 5 6 7 8 9 10

Through the analysis of the initial three rankings, it was determined that the mouths of the

Gavaraget, Masrik, and Martuni rivers were locations most in need of water treatment while still able to harbor and optimize duckweed biomass yield (Figure 6). 46

After ranking the ten rivers and selecting the Gavaraget, Masrik, and Martuni as rivers of interest,

Google Maps was used to make a list of various towns and villages situated around the rivers.

Topographic data for the area were utilized, revealing that the elevation in each area around the towns is relatively constant, only varying at most by about 150 m (500 ft) in many towns. In turn, there should not be any logistical issues due to elevation when farming or transporting duckweed. Alongside the topographic data for each town, the most recent population data from the 2011 Population Census of the

Republic of Armenia published in 2013 were utilized to approximate the size of the towns within the regions of interest (Table 4). Eighteen towns were selected for conducting potential interviews.

The final deliverable from this analysis was the hydrology data calculated for each of the three rivers. As shown in Table 5, the Gavaraget river has the largest contribution of nitrogen and phosphorus,

179,662 kg (396,087 lbs) and 29,619 kg (65,299 lbs) respectively, flowing into Lake Sevan, followed by the Masrik and the Martuni rivers. Because of this, it is recommended that the Gavaraget river be focused on for the initial implementation so that the greatest impact can occur before further development begins. 47

Table 4: Municipality Population Data Circa 2011. Towns marked in yellow are the locations where interviews were conducted. (*Population data obtained from Google.com) Gavaraget Municipalities Population Gegharkunik 1,654* Sarukhan 8,309 Lanjaghbyur 2,115 Artsvaqar N/A Karmirgyugh 4,964 Gavar 20,765 Noratus 6,732 Gandzak 3,815 Masrik Municipalities Population Torfavan 490 Vanevan 367 Lusakunq 1,440 Khachaghbyur 1,064 Karchaghbyur 2,337 Vardenis 12,685 Martuni Municipalities Population Martuni 12,894 Tsakqar 2,692 Tsovasar 2,708 Dzoragyugh 4,737

Table 5: Hydrology Data Calculated for the Gavaraget, Martuni, and Masrik Rivers Gavaraget Average Nitrogen Average Phosphorus Average Volumetric Total Yearly Volume Yearly Total Yearly Total Concentration Concentration Flow Rate (m³/s) (m³) Nitrogen (kg) Phosphorus (kg) (mg/L) (mg/L) 2.65 1.10 3.48 1.10E+08 292530 111047 Martuni Average Nitrogen Average Phosphorus Average Volumetric Total Yearly Volume Yearly Total Yearly Total Concentration Concentration Flow Rate (m³/s) (m³) Nitrogen (kg) Phosphorus (kg) (mg/L) (mg/L) 2.23 0.48 1.72 5.42E+07 179662 29619 Masrik Average Nitrogen Average Phosphorus Average Volumetric Total Yearly Volume Yearly Total Yearly Total Concentration Concentration Flow Rate (m³/s) (m³) Nitrogen (kg) Phosphorus (kg) (mg/L) (mg/L) 2.19 0.32 3.24 1.02E+08 214605 32768 *Nutrient Concentration Data for February in the Martuni River was not available, requiring an estimation 48

Objective 2. Assessment of Local Attitudes

To better understand local attitudes towards the implementation of a duckweed phytoremediation system and estimate the support and infrastructure available for farm implementation, representatives from the previously identified municipalities surrounding Gavaraget, Masrik, and Martuni were interviewed by Hovhannes. The first group of interviews conducted took place in the Gavar region, by the western shore of Lake Sevan. The Sarukhan village was found to be a large community with inflows of sewage and agricultural waste into Gavaraget. This, as well as there being a fishery located in the river

(Figure 7), makes the town a potential location for duckweed implementation. Issues arose with local attitudes towards the project however. Although the Sarukhan municipality representative agreed that the water quality of Lake Sevan is an issue, he did not find the problem of a higher priority for the town due to the municipality’s resources being focused on other local initiatives. Also, the fishery owner had no interest in the project, nor did he believe that the fishery was responsible for waste inflow or that nutrient concentrations were the cause of eutrophication. Finally, due to the carnivorous nature of his fish, the owner believed he would have no use for a duckweed-based fish feed. 49

The second village visited in the Gavar region was Lanjaghbyur. The sentiments in this community were similar to those encountered in Sarukhan. The Lanjaghbyur representative understood the goals of the project and was interested in the economic opportunities presented by duckweed farming, but was initially hesitant to implement a duckweed farm in Lanjaghbyur. He believed that the AUA

Acopian Center would have trouble convincing community members to volunteer for or commit resources towards the initial farm construction, and also raised concerns that there would be limited machinery available for usage as well. He was also concerned that there was a possibility of the Gavaraget River flooding, which may present a danger to the local populace and deter involvement. Finally, a possible issue that may occur with the maintenance of duckweed farms is the lack of volunteers during times when a large part of the population goes abroad for work. In order to expand upon the concern raised by the

Lanjaghbyur representative, another local community member who owned a livestock farm was contacted. He was more optimistic regarding the feasibility of duckweed implementation, as his farm had contemporary infrastructure built in the area and he believed that the community would have a favorable view towards modern approaches of animal feed production. His main concerns were similar to the municipality representative though, as he felt financing would be an issue and he had little faith in government or non-government organizations implementing projects in the municipality. He was interested and willing to invest in the possibility of applying post-harvest duckweed biomass as animal feed, but only if it were tested by other farms similar to his own beforehand. It was found that, currently, he can produce 7 kg (15.4 lbs) of animal feed himself from 1 kg (2.2 lbs) of barley; therefore an alternative method would need to be more efficient and cost-effective.

The third municipality visited in the Gavar region was the village of Gegharkunik. The

Gegharkunik municipality representative was very cooperative and supportive of the project’s goal of farm implementation, strongly agreeing that the water quality of Lake Sevan is a pressing issue that must be addressed. He highlighted topographical features within the town as well, citing that the village is located in a mountainous region and that the Gavaraget River is fast-flowing due to the slope. Because of this location, there is an issue regarding available land for duckweed farms. He also raised concerns about 50 his village’s ability to cooperate with other neighboring municipalities. Besides these issues, the town presents aspects that make it a potential location for implementation. He explained that he was one of the largest farmers in the municipality, and was willing to buy animal feed produced from duckweed harvest, so long as there is a method of preserving the feed for an extended period of time. He also revealed that the local school has an environmental protection club, which may be interested in working with the project. The final recommendation was that further efforts should be made to explore the river cleaning station found near the Artsvaqar and Noratus communities. It was explained that this location may be suitable for duckweed pilot farm development.

The fourth and final village of the Gavar region that was visited was Artsvaqar. This location was found to be promising as it is home to an ecotourism pond that had previously grown duckweed naturally

(Figure 8). This pond was also already connected to the Gavaraget River through a tunnel connection.

Yet, due to the community's focus on ecotourism rather than economic gain from agriculture, the previously grown duckweed was removed. There was found to be a large population of ducks inhabiting the area as well, presenting a possible risk towards the newly implemented duckweed as ducks consume duckweed, decreasing the amount of biomass that can be produced. The ducks may also transport duckweed from the farms to other waters where it can grow without restriction. These ducks were important in the municipality's ecotourism efforts, meaning that they would not be removed from the area.

The owner of the local pond was hesitant to support any duckweed growth initiatives due to past struggles with the previously growing duckweed. He did express interest though, and was willing to support the project if its economic benefits were researched further. 51

The next region in which interviews were conducted was Martuni. The first municipality that was contacted was Tsakqar, in which a representative was interviewed. The community representative was found to be rather cooperative, willing to maintain consistent communication in the future as the project progresses. Yet issues soon were revealed, as the town did not have a large population of residents working in agricultural industries, nor did it have land or resources that it was willing to donate towards project efforts. The village also did not seem as it would benefit from duckweed applications, as there was little agricultural development nearby. Outside of these issues, there were concerns regarding where duckweed seeds would be supplied from and how they would be acquired, which must be considered in the future.

A representative from Tsovasar, the next village, was contacted. He expressed that there was no workforce available for farm implementation and management, nor were there interested stakeholders willing to invest resources in the project. These same sentiments were shared by the municipality representative of Dzoragyugjh. Based on these two interviews, it would seem that there was little interest in implementing a duckweed farm. We assume this may be influenced by other initiatives occurring in the region, such as the Gesellschaft für Internationale Zusammenarbeit (GIZ) initiative taking place in the neighboring village of Vardenik. GIZ provides cost-efficient and effective services for sustainable 52 development, and has been working within Armenia since the early 1990s. This internationally sponsored project has a similar objective to our own, seeking to implement a water treatment system in the surrounding area. Therefore, it would be redundant for the community to support both initiatives.

The third and final set of interviews was conducted in the Vardenis region, located around the

Karchaghbyur and Masrik Rivers. The first town visited was Karchaghbyur located along the

Karchaghbyur River (Figure 9) and about 12 km (7.5 mi) from the Martuni River. The Karchaghbyur municipality was proactive and seemed very supportive of the duckweed farming project, having access to open land alongside the river and the workforce needed to construct and maintain duckweed farms. The municipality representative expressed interest in founding a poultry farm if duckweed farms are implemented and could financially benefit the town. Lake Sevan National Park operates fisheries nearby breeding Sevan trout that could present a potential consumer of harvested duckweed. Aside from the

National Park Fisheries and possible implementation of a poultry farm, there are no other large farms nearby, raising concerns about the benefits local farms would receive for assisting in the project alongside concerns about funding for the project. A second interview was conducted with the owner of private fish ponds. The pond owner has access to three small ponds connected to the Karchaghbyur River that are currently used for raising a small number of fish and generating hydroelectric power. He expressed that he has the time as well as some assistance available to maintain the duckweed farms; however, he would only consider supporting the project if it was guaranteed to be a profitable venture and wouldn’t require any monetary investment on his behalf. 53

The second town visited in the Vardenis region was Khachaghbyur. The representative noted that the workforce does not reside in the village and the town itself does not own land that could be used for duckweed farms. He also expressed that the village does not have the funds to support financing the project and has no experience with initiatives of this nature. Within the town, there are no large farms that could make use of harvested duckweed and only a single fishery, which is currently not raising fish. This is unlikely to change, but the town representative was going to pass the information about the duckweed project to possible stakeholders.

The third and final town visited in the Vardenis region was Lusakunq. The town is interested in sustainable development and has organized garbage collections in efforts to reduce pollution. Pollution and water quality seem to be issues of high priority within the village. It was suggested that prior to duckweed farms being implemented in the region, efforts should be made to clean up litter along the banks of the river. Despite the initial support of Lusakunq, several concerns were raised. The first was that the land around the riverbanks is inaccessible to heavy machinery and is privately owned. The town also 54 cannot commit resources to support the project. In addition to concerns about resources, the river’s water was said to be very cold which could negatively impact the growth of duckweed, and another environmental project in Lusakunq focused on tree planting was unsuccessful. The town representative is in contact with the regional government and may discuss the implementation of duckweed farms, possibly leading to government support. An owner of a local fishery was interviewed as well. The owner had previously tried to cultivate fish in the river-fed basin but ran out of funding and does not plan to use the basin in at least the next five years. This opens up the possibility of implementing a duckweed farm without developing new infrastructure, but the fishery owner is hesitant to invest more of their time and money on any type of cultivation. 55

Objective 3. Opportunity Analysis of Duckweed Farming

To determine the potential economic benefit of duckweed harvesting, an algorithm was developed to estimate the biomass yield from duckweed farming. This model used a variety of parameters that affect duckweed growth as inputs, including water temperature, photoperiod, pH, and nutrient concentration, in order to apply the procedure in a variety of environments. Using EUWI+ data collected from the three rivers in 2019, biomass calculations were carried out over a time period of eight months, from March to

October. The months outside of this range were not considered as the mean water temperatures for each river during this month was below the 5℃ necessary for duckweed growth. The initial calculation estimated the intrinsic growth rate based on temperature, photoperiod, pH, and nutrient concentration data. Intrinsic growth rate is considered the reproduction rate minus the death rate. The next calculation found the specific growth rate, or the increase in biomass per unit of concentration. This calculation allowed for the mat density of the duckweed harvested to be determined based upon the harvest period and initial mat density. These equations are shown in Figure 10 below.

The complete spreadsheet of calculations developed was provided to the sponsor for further use.

It was estimated that a total of 1,093 kg/m² (223.864 lbs/ft²) of dry duckweed biomass could be harvested from the Gavaraget River. For the Martuni and Masrik rivers, this biomass value was estimated to be 755 kg/m² (154.636 lbs/ft²) and 774 kg/m² (158.528 lbs/ft²) respectively. 56

Table 6. Biomass Estimations for the Months Between March and October

Gavaraget Masrik Martuni

Dry Biomass Dry Biomass Dry Biomass Month Month Month (kg/m²) (kg/m²) (kg/m²)

March 6.1 March 4.4 March 4.0

April 14.6 April 11.9 April 7.8

May 76.5 May 51.3 May 27.1

June 182.3 June 108.5 June 228.5

July 254.9 July 205.9 July 120.6

August 263.8 August 176.5 August 187.8

September 198.0 September 116.9 September 133.6

October 97.2 October 98.4 October 45.3

TOTAL (kg/m²) 1,093.4 TOTAL (kg) 773.8 TOTAL (kg) 754.8

Combined Dry Total 2,622.0 (kg/m²)

Combined Dry Total 537.0 (lbs/ft²)

A study in Bangladesh conducted at the Department of Fisheries Management tested the impact duckweed has when used as a supplement for fish-feed. Five species of fish were used in this experiment: silver carp, mrigal (white carp), sharputi (silver barb), tilapia, and common carp. In all five species, it was seen that over a 90-day period, fish fed with duckweed grew to a larger size than their counterparts not fed duckweed. The mrigal had the smallest difference in size with only a 2.17% size increase in the fish fed with duckweed compared to those not fed with duckweed. The tilapia had the largest difference in size with an 87.85% size increase in the fish fed with duckweed compared to the tilapia not fed with duckweed supplementing the commercial fish-feed. Across the five species the average size increase was 31.2%.

This increase in size of the fish fed with duckweed is likely a result of dried duckweed having a high protein content and a low fiber content which is ideal for the nutritional needs of fish (Kabir, 2009). A 57 similar increase in size of animals when fed duckweed was also seen in chickens when used as poultry feed (Haustein et al., 1992) as well as in ruminants when Holstein cattle were fed a mix of maize silage and duckweed (as opposed to just maize silage). Despite the size increase noted in the Holstein cattle, research has suggested that although duckweed is able to provide ruminants with essential nutrients, it is unable to provide the cattle with significant amounts of protein as duckweed protein is easily fermentable and cannot be absorbed by the ruminants. This is due to the multi-stomach nature of ruminants, the protein in the duckweed would ferment and be absorbed by the bacteria within the stomach of the ruminant before the ruminant itself can absorb the protein (Leng, 1995).

While studies have shown that duckweed has several benefits when used as feed for fish and other animals, the amount of duckweed used as feed and when the duckweed is fed to the animals varies depending on the animal. Looking first at poultry, various studies note that chicks have stunted growth when fed duckweed, but older birds were seen to have higher growth rates than their counterparts fed with commercial feed. Broiler and layer chickens also seem to have their growth affected differently by duckweed with layer chickens seeing fewer negative impacts on growth, likely because layer chickens aren’t grown to as big a size as broiler chickens. No negative impacts were seen in the egg quality of layer chickens when fed with 25% duckweed (Haustein et al., 1992; Haustein et al., 1988). When used as fish-feed, various studies noted negative impacts on fish growth when too much of the commercial diet was replaced with duckweed. These negative impacts included decreased survival rates of the fish and impaired growth rates. Despite several studies noting negative effects when fish are fed too much duckweed, the amount of duckweed able to be fed to fish without seeing negative impacts varies study to study and seems to be dependent in part on the species of fish in the studies. One study found that 3-5% of the total fish body weight replaced by duckweed as feed was optimal (Hassan & Edwards, 1992), while other studies saw that omnivorous fish such as rohu (a type of Asian carp) saw significant growth increases when up to 30% of the commercial feed was replaced by duckweed (Anthonius et al. 2016). A third study showed duckweed increased the growth rate of several species of fish when the fish were fed duckweed totaling 20% of the fishes’ total weight (Kabir, 2009). 58

Research has also shown duckweed to be a promising standalone fertilizer and fertilizer amendment. Experiments conducted at Pennsylvania State University found that duckweed when used as a fertilizer is likely to have less nitrogen runoff compared to commercial fertilizer. In addition, duckweed increased the nutritional value of sorghum grown with duckweed as fertilizer. Research also shows that duckweed helps soil retain other nutrients such as carbon and calcium. Overall, duckweed performed similarly to commercial fertilizer while providing better nutritional value for plants with the drawback that duckweed significantly impaired germination by 50% which would lead to reduced crop yields of

32% compared to commercial fertilizer. Further tests suggest that duckweed itself does not necessarily inhibit germination, but rather other factors such as the growing conditions of duckweed result in the duckweed causing lack of germination. In the case of the Pennsylvania State study, high concentrations of salt in the duckweed caused by the growing duckweed in water containing high salt levels stunted germination (Pulido, 2016), something that can be avoided if duckweed is grown in water with low salt concentrations.

While duckweed offers many benefits and a few drawbacks when used as feed and fertilizer, the duckweed must be tested for heavy metal contamination prior to use as feed or fertilizer to prevent even greater negative effects. If contaminated duckweed is fed to animals the animals could suffer from heavy metal poisoning or the heavy metals could be transferred to humans eating meat or crops contaminated with heavy metal. One study found that L. gibba heavy metals affected the levels of chlorophyll and carotenoid in the duckweed. Chlorophyll a levels decreased, chlorophyll b levels increased, and carotenoids levels surpassed that of chlorophyll (a+b). These changes in pigment levels change the physical appearance of the duckweed leaves as they undergo chlorosis, changing from a dark green to a light green to a whitish hue respective of the levels of the heavy metal. Despite the appearance change in leaves offering a general idea of heavy metal contamination levels, spectroscopy is the only way to determine the actual levels of metal contamination (Hegazy et al., 2009).

Bioethanol research revealed that using duckweed as a feed source for bioethanol production has many benefits. First, maize is currently the main feed source for bioethanol production. By using 59 duckweed, the stress of excessive demand created from maize-based bioethanol production placed upon the corn industry is reduced, as corn is also an important food and feed source for livestock. Intensive corn production also has high requirements for agricultural inputs, which results in substantial pollution and soil erosion (Cheng et al., 2009). Duckweed not only has a higher growth rate and longer growth period than traditional field crops, but also has less environmental impacts as it requires limited agricultural inputs and helps to treat wastewater. The main method required to produce ethanol from a duckweed-based feed source is through enzymatic hydrolysis for fermentable sugar production. This byproduct is then fermented anaerobically by yeast to produce ethanol (Cui et al., 2015).

In order to produce the largest yield of ethanol possible, efforts must be made to increase the starch content of the harvested duckweed. One method that was researched that may prove applicable for our project is nutrient starvation. One study conducted by Cheng et al., (2011) introduced 3 kg (6.61 lbs) of fresh duckweed into well water in order to induce nutrient starvation. After ten days of this process, the starch content obtained from the biomass was increased by 198.6%. By introducing various levels of

NaCl to increase salinity, this content increased to 214.8%. Because of these findings, it is recommended that this process is used if bioethanol production is carried out using the harvested duckweed. The low temperatures present in Armenia during the spring and fall will also help in this process, as low temperatures favor starch development due to the reduced respiration rates at night. Cui et al. (2011) reported that Spirodela polyrhiza grown at temperatures around 5℃ (41°F) had a starch content 114% higher than that grown at 25℃ (77°F).

Once methods to create a high starch biomass were recorded, the procedure by which ethanol is produced was considered. Based upon the procedure outlined by Cheng et al. (2014), enzymatic hydrolysis and yeast fermentation can be conducted in 14 L continuous stirred tank reactors (CTSR).

Their procedure had suggested that the dry duckweed be ground before the process begins, yet this could be an optional step in order to reduce infrastructure cost. Beyond this, the initial hydrolysate mixture was comprised of 1 kg of duckweed dry biomass, 4 L of MOPS buffer (pH 7), and 1.5 x 10⁶ units of alpha-amylase. This slurry was incubated at 90℃ (194°F) for 45 minutes and an agitator rotation speed of 60

150 rpm was applied throughout hydrolysis. Two more additives, Glacial acetic acid to adjust pH to 4.5 and 2 x 10⁵ units of pullulanase, were introduced. The incubation continued at 60℃ (140°F) for another

30 minutes. Temperature was then decreased to 50℃ (122°F) and 10⁵ units of amyloglucosidase was added so that the mixture could be once again incubated at that temperature for four more hours. Once complete, the fermentation process began by adding 2 N (Normality) NaOH while fermenting with yeast at 30℃ (86°F) for 72 hours.

Through this process, it was reported that the reduced sugar recovery reached 96.8% after hydrolysis and ethanol production was 97.8% of the theoretical yield. This was an overall conversion rate of 94.7%. Based upon this process, 3.35 kg/m² yielded 0.942 kg/m² of starch (31% content) and 0.642

L/m² of ethanol, about 50% higher than that obtained using the same biomass of maize. Other experiments carried out by Cheng et al. (2009) found that ethanol yield was 258 mg per gram of dry duckweed biomass at a starch content of 45.8%. These results indicate that, under the proper conditions dry, duckweed biomass can produce significant quantities of starch, used as a source to produce ethanol.

While the exact amount of ethanol that can be produced will vary depending on the duckweed species used and the starch content of the duckweed, one study found that between 0.16 and 0.19 grams ethanol/g dry mass of Landoltia punctata, Wolffia arrhiza, S. polyrhiza, and Lemna aequinoctialis can be produced

(Faizal et al., 2021). Based on these data and our estimated biomass yield, it is estimated (Appendix D) that a single duckweed farm could produce between 153 L (40.42 gallons) and 357.42 L (94.42 gallons) of ethanol. From this, it can be estimated that the total value of the ethanol created would be between

$94.86 and $221.60 based upon a bioethanol price of $0.62 per liter reported by the U.S. Grains Council for the week of May 5th, 2021. 61

Objective 4. Design of a Duckweed Pilot Farm

Initial work on our fourth objective was focused on selecting which duckweed species to use in the pilot farm(s) so that design considerations could be made around any particular needs of the duckweed species grown in the farm. Analysis of Landolt’s The family of Lemnaceae- a monographic study Volume

I, which studied 14 species of duckweed [L. minor, L. gibba, L. trisulca, L. punctata (formerly Spirodela punctata), Lemna turionifera, Lemna minuscula, S. polyrhiza, Lemna valdiviana, Lemna minuscula,

Wolffiella gladiata, L. aequinoctialis, Wolffia globosa, W. arrhiza, and Wolffia columbiana] revealed that several species of duckweed are capable of overwintering and successfully regrowing in the spring.

During the winter of 1957-58 in Zürich, basins in which the duckweed was growing had a layer of ice 30 cm (1 foot) thick, with a mean water temperature of -8.7℃ (16.3°F) for several weeks. Once the ice melted in the spring, S. polyrhiza, L. gibba, L. turionifera, L. minor and L. trisulca successfully came back with a high survival rate, while L. punctata and L. minuscula also regrew with a low survival rate.

All remaining duckweed species did not regrow in the spring. It is important to note that while half of the species did regrow in the spring, L. minor, L. gibba, and L. trisucla regrew about three weeks earlier than the other four species while S. polyrhiza began to regrow six weeks later. Other studies support these observations, stating that L. minor is generally the first species to regrow in the spring with turion-forming species regrowing a few weeks later. Turion-forming species develop later in the season as the starch-filled buds called turions formed for overwintering require a prolonged exposure to warmer temperatures to grow into mature duckweed.

Looking further at the seven species that survived the 1957-58 Swiss winter, L. minor, L. gibba, and L. trisulca were all found to be able to grow at 4–5℃(39–41°F), L. minuscula, L. turionifera, and L. punctata were all found to be able to grow at 6–9℃ (43–48°F), and S. polyrhiza was found to be able to grow a 9–13.5℃ (48–56°F) (Landolt, 1986). Based on the minimum growth temperature, L. minor, L. gibba, and L. trisulca were selected to be recommended for use in the duckweed pilot farms. Looking at the protein and starch levels of these three species, L. gibba has the highest levels at ~28% the dry weight 62 of duckweed and ~7% the dry weight of duckweed respectively. L. minor has a protein level at ~25% the dry weight of duckweed and ~4% the dry weight of duckweed (Appenroth et al., 2018). L. trisulca had a protein content ranging between 7.1 and 17% of the dry weight of duckweed (Meyers, 1982). The starch content of L. trisulca could not be found in literature.

With the selection of three potential duckweed species to use in the phytoremediation system, the next step is to select whether to grow the duckweed as a monoculture or a polyculture. Two Chinese studies (Zhao et al., 2014; Li et al., 2016) researched the impact duckweed monocultures and polycultures have on the growth rate, starch content, and protein content of various duckweed species when influenced by varying temperatures, light conditions, and phosphorus and nitrogen levels. In the 2014 study, researchers first looked at L. minor and L. punctata grown as monocultures and as a two-species polyculture under varying conditions. Grown at three different temperatures [20, 25, and 30℃ (68, 77, and 86°F)] it was found that the polyculture showed no distinct advantages or disadvantages compared to the monocultures in regards to biomass production or the starch content of the duckweed. Under varying light levels, it was found that the polyculture promoted the growth rate and increased the starch content of duckweed under low light levels (2000 and 5000 lux). Under varying concentrations of phosphorus and nitrogen, the polyculture had a growth rate between that of both monocultures; however, it was noted that the duckweed grown in polyculture had longer roots than their respective monocultures. Extended root systems have been shown to be capable of increasing the uptake of nutrients, particularly phosphorus suggesting that the polyculture of duckweed offers some advantages in regards to nutrient uptake. Low levels of phosphorus and nitrogen were shown to increase the starch content of duckweed, while the polyculture also promoted starch accumulation at low nutrient levels. Other tests looked at the absorption rate of ammonium and phosphate from the swine wastewater by the duckweed. The polyculture had the highest rate of ammonium uptake three days after initial culturing, and after six days the polyculture had the highest rate of phosphate uptake. The rate of ammonium and phosphate uptake in the wastewater was also measured in monocultures and polycultures of L. minor, L. punctata, and S. polyrhiza. Ammonium uptake in a two-species polyculture was greater than any of the three monocultures and the three-species 63 polyculture. Phosphate uptake was better in two- and three-species polycultures than it was in any monoculture (Zhao et al., 2014).

To build of these results, the second study conducted in 2016 was performed to see how the growth rate, starch content, and protein content of L. aequinoctialis, L. punctata, and S. polyrhiza were impacted by being grown as monocultures and polycultures under varying environmental conditions.

Consistent with the 2014 study, it was found that there was no significant difference in the biomass production and starch content in the monocultures and polycultures at varying temperatures; however, at low temperatures [20℃ (68°F)], polycultures of all three species as well as a polycultures of L. aequinoctialis and S. polyrhiza yielded higher protein levels than the monocultures did. Under varying light intensities, it was found that using monoculture or polyculture has no significant impact on duckweed growth rate. The starch content under varying light levels showed that the three species polyculture had higher starch content than the monocultures at low light intensity and the L. punctata and

L. aequinoctialis had higher starch content at high light intensity compared to their monocultures. These results conflict with those of the 2014 study, but researchers noted that the duckweed used in the previous study was collected from a region that is often cloudy while the duckweed used in the more recent study was collected from a region that receives more sunlight. This suggests that growth rate and composition of the duckweed is impacted in part by adaptations to the specific environment that plants grow. In regards to protein content at varying light levels it was seen that polycultures had higher protein levels under all light conditions. Under varying concentrations of phosphorus and nitrogen, no significant difference in growth rate between monoculture and polyculture was seen. At low nitrogen (N) and phosphorus (P) levels (ranging from 3.5 mg·N/L to 0 mg·N/L and 1.5 mg·P/L to 0 mg·P/L, respectively) polycultures increased both starch and protein levels with no significant differences seen in the starch and protein levels between monocultures and polycultures when the duckweed were grown under nutrient-rich conditions (35 mg·N/L and 15 mg·P/L) (Li et al., 2016).

After determining L. gibba, L. minor, and L. trisulca as the species to be recommended, the design of the duckweed farms can be made allowing for any species-specific accommodations that may 64 be needed. The building of the farm will be dependent on the resources available. Due to these restrictions, we will begin by outlining general specifications of the farm. The shape of the pond should be narrow and long, as opposed to wide and box shaped. If the width of the pond is over 6 m (20 ft), the harvesting process will be more difficult. If the land available is not long and narrow, the pond should be an “s” shape to increase the length of the coastline. The depth of the pond should be between 1 to 1.5 m (3 to 5 ft) to ensure that the duckweed is easily accessible for harvest. Ideally, the pond will be installed on flat land to limit the resources required for installation while the water should be still and in an area with minimal wind exposure. Systems must be in place to ensure a water of slow current enters the pond and duckweed stays within the farm boundaries during water outflow. Once the location has been chosen, the water entry and runoff systems need to be installed. Figure 11 is a sample design of a very simple and low-cost variant of the pond design. 65

Water enters and exits the pond through small canals in order to fill the pond. The size of the entry canal is dependent on the current of the river, as well as the equipment used to dig the canal. The exit for the canal has a similar design and will be located downstream of the entry canal. The mouth of the canal has a wire mesh or net attached that will prevent flora from exiting in the runoff, keeping the duckweed within the farm boundaries. The wire mesh and rod will be a cost effective option for farm containment. Other costs associated with implementation will be in securing duckweed and the manual labor needed to dig the pond if volunteers are not found. Without proper equipment and community support, it may not be feasible to pay for that much manual labor. The transfer of duckweed into the pond should be inexpensive due to the recommended species being native to Armenia.

Once the duckweed farm has been built, the only continual investment into the farm will be the harvesting and drying of the duckweed on a regular basis. This will require an investment of both time and money; however, the amount required of each greatly depends on the harvesting methods and if manual or automated methods are utilized. Two methods of manual harvesting include using a pond rake or a PondSkim (Figure 12). Both of these tools are cheap and reusable, with pond rakes ranging in cost from

$99.99 to $249.99 while a Pondskim costs $149.99 (The Pond Guy, n.d).

While these tools require little initial investment, they are time consuming and impractical for large scale duckweed harvests. Yet, given the size of the pilot pond design, this shouldn’t pose a problem.

Considerations must also be made about who will manually be harvesting the duckweed as it requires a high degree of physical labor. Volunteers will be needed to harvest the duckweed and perform routine 66 maintenance. If there isn't enough support to do so, employees would need to be hired, adding to the costs of maintaining the duckweed farm. Automated harvesting, unlike manual methods, has very high initial investment costs in exchange for faster harvesting capabilities and minimal human involvement. One automated collection method is The ProSkimmer System manufactured and sold by ProSkim. The

ProSkimmer System floats on the surface of the water and only intakes the weeds floating on the surface.

The weeds are then pumped into an onshore collection chamber before the water is returned back to the lake or pond. At a price of $6,950, The ProSkimmer System requires minimal maintenance after installation to empty the collection chamber (Proskim duckweed and watermeal removal, n.d.). Other automated methods exist such as using aquatic harvesters similar to what Aquamarine offers; however, these harvesters are very expensive ranging between $69,980 to $199,980 and are impractical for use in a small duckweed pond such as the pilot farm. This is due to both the size of the machine and the cost to purchase one (Aquamarine- aquatic weed harvesting equipment, n.d).

Once the duckweed has been harvested, the duckweed must then be dried for post-product use, unless the duckweed is being fed to livestock other than poultry or fish or is being used as fertilizer. In these cases, the drying process would be optional. There are two methods of drying harvested duckweed: solar drying and oven drying. Solar drying is the cheapest drying method only requiring black plastic sheets and sunlight. As the duckweed is being harvested it can be placed directly onto the plastic sheets and left under the sun for several days. These sheets must be covered if it rains during the drying period in order to prevent complications. Large quantities of duckweed can be dried this way, only limited by the area covered by the plastic (Leng, 1999). Oven drying, in comparison to solar drying, is a much faster process. Based on methods presented in literature, this method of drying would allow duckweed to be dried at 105°C (221°F) for 24 hours (Laube & Wohler, 1973) or at 40°C (104°F) for 30 hours (van den

Berg et al., 2015). Despite the advantage of drying duckweed faster, oven drying is a more costly process than solar drying due to the energy demand of running an oven for hours on end. In addition, simple home ovens are only large enough to dry small batches of duckweed. Large harvests of duckweed would require industrial ovens to dry the duckweed in a timely manner in order to prevent the duckweed from rotting. 67

Recommendations

Farm Location Recommendation

During interviews with representatives of several municipalities, concerns were raised regarding the availability of land for duckweed farm development. In order to ensure that a pilot farm can be established and further development can be carried out, we recommend that the AUA Acopian Center for the Environment determines a process to identify open land or water for pilot sites in the regions surrounding Gavaraget, Masrik, and Martuni. Land or water in regions in which there was initial support for the project, such as Karchaghbyur, should be considered primary locations for consideration. It is also important to consider the hydrology data which was calculated, in which the Gavaraget river had the largest outflow of nutrients. Because of this, the Gavar region should be the initial location in which implementation is attempted. These locations must also be accessible and able to support the pilot farm design. Finally, we recommend that resources are committed to identifying possible legal and logistical issues that may arise during initial development and continued maintenance of duckweed farms.

Duckweed Post-Harvest Application Recommendation

Although we were able to utilize an algorithm to estimate the growth rate and biomass output in conditions found in the Gavaraget, Masrik, and Martuni rivers, pilot farm testing is needed to corroborate our calculations. Through controlled growth in water collected from possible farm locations, further calculations can be made to determine the economic feasibility of implementing duckweed farms. This testing would also determine the maximum mat density, or maximum density of duckweed on the water surface before there is a detrimental effect on growth rate, before harvest as well as the harvest period needed for consistent maintenance and optimal biomass output. It would also be important to find the starch and protein content of the various species of duckweed when grown in water sourced from the various rivers. This would help to predict the effectiveness of post-harvest applications and determine if there is a need for added infrastructure for processes like nutrient starvation and salinity treatment. 68

Through this, the outlined post-harvest biomass applications can be produced and tested in order to compare them to these predictions and their current alternatives, such as maize-based bioethanol and soybean-based animal feed. After this process is completed, a more thorough understanding of the estimated economic impact can be created and pitched to possible volunteers and interested stakeholders in the municipalities of interest. Further data collection is also recommended in order to increase the accuracy of the model, as water temperature and photoperiod data was not collected from the tributaries directly.

Literature analysis revealed a disparity among the duckweed species with regards to minimum growth temperature, protein content, starch content, growth rate, and many other properties. The Lake

Sevan Basin is one of the cooler regions of Armenia, due to its high elevation. As a result, in order to maximize the timeframe in which duckweed is able to grow, the minimum growth temperature of duckweed species was prioritized in selecting which species to use in the duckweed farms. We recommend using L. minor and L. gibba in the duckweed farms with possible inclusion of L. trisulca as the next best option. All three species have been documented to grow at temperatures as low as 4–5°C

(39–41°F), which is 1–5°C (2-9°F) colder than any other documented species, and would be the first to regrow in the spring, further maximizing the timeframe of potential duckweed growth. L. minor and L. gibba were selected over L. trisulca as studies have shown L. triscula has a significantly lower protein content than the other two species, making it a less ideal option for use as animal feed post-harvest. L. minor is also recommended in part because it is native to Armenia. This simplifies the replanting process, as the duckweed will be readily available in the region. Despite L. trisulca being a less ideal option for post-harvest applications, use of L. trisulca alongside L. minor and L. gibba can potentially offer benefits if grown in a polyculture. While literature could not be found regarding the impact of polyculture specifically using L. trisulca, L. minor, and L. gibba, research has demonstrated that polycultures of two or three species of duckweed may have more benefits than monocultures. Under the phosphorus and nitrogen levels most similar to those in the rivers in which the farms will be implemented, polycultures of duckweed appear to increase the starch and protein content of the duckweed under certain growing 69 conditions, including those most similar to what will be encountered in the Lake Sevan Basin. This will improve the quality of the post-harvest product if duckweed is used as animal feed or biofuel production.

While we recommend that polycultures are used, the species used in the polyculture may impact its effects on changing the protein and starch content of duckweed, so testing of protein content and starch content of the duckweed is recommended under different polyculture conditions such as all three species grown together or just L. minor and L. gibba grown together.

Interviews with municipality representatives revealed that many towns in the Lake Sevan Basin have agriculture or animal husbandry as large components of their economies. Based on the nature of the local economies, we recommend that harvested duckweed is used for animal feed supplements, compost, and fertilizer supplements. Aside from the initial investment of creating the duckweed farms, farmers would have the ability to produce animal feed and fertilizer at little to no cost to them. On top of this, duckweed-supplemented animal feed has been shown to increase the size of farm-raised fish, poultry, and cattle, further benefiting farmers by raising the market value of their livestock. Duckweed would also make a good amendment to fertilizer as duckweed is rich in phosphorus and nitrogen, decomposes quickly, and can be used when wet or dry. Soil treated with duckweed has been shown to have increased phosphorus and nitrogen levels as well as decreased levels of ammonia and nitrate leaching when both treated with duckweed as fertilizer and when duckweed was mixed with commercial fertilizer, improving soil quality.

Since research into optimizing duckweed as animal and fish-feed research has varying conclusions, we recommend further research looking into the maximum amount of duckweed can be used to replace commercial feed before the negative impacts are seen in animals. Conclusions about using duckweed as poultry feed are more consistent than conclusions made about using duckweed as animal feed, so as a starting point, we recommend that for poultry-feed for mature birds, dry duckweed is used to minimize the water content and maximize the protein content taken up by the birds. Duckweed should not be used for chicks unless more research is done to avoid stunting the growth of the chicks. We recommend that duckweed replace 25% of commercial feed up to a maximum of 40% replacement of 70 commercial feed for adult chickens, with the option to replace a lesser percentage at the discretion of poultry farmers. If duckweed is used for a feed supplement for poultry other than chickens, we recommended starting with less than 25% commercial feed replacement and monitoring the birds’ weight and health for any negative effects before increasing the amount of duckweed used in the feed. Guidelines based on literature for using duckweed as a supplement for fish-feed is less clear. We recommend that initial inclusion of duckweed as a feed supplement for fish begins by replacing a small amount, about 5% of the fishes’ total body weight with duckweed instead of commercial feed, with continual monitoring of the fish for any adverse effects following each increase. After monitoring the growth and health of the fish for any negative side effects for several weeks, the amount of duckweed can be increased by another

5% up to a maximum of 30% of the fish’s body weight for omnivorous and herbivorous fish.

If duckweed is being used as fertilizer, we recommend that it is only used for crops that have passed the germination stage until further small-scale tests are performed to ensure that duckweed will not inhibit the germination of crops when used as fertilizer. To use duckweed as fertilizer, there are three recommended methods: drying the duckweed and using it directly, drying the duckweed and mixing it with commercial fertilizer if enough biomass cannot be produced for duckweed alone to be used as fertilizer, or adding duckweed to a compost pile after it has been harvested. In the event that an excess of duckweed is produced, the duckweed should be dried, placed in airtight containers, and stored away from sunlight to prevent degradation of the duckweed before it can be used as fertilizer or animal feed in the future.

Through our literature review of the viable applications of duckweed biomass, it was found that the production of bioethanol was feasible when using high starch content duckweed as a feedstock.

Although the suboptimal phosphorus and nitrogen levels in the Lake Sevan’s tributaries would create a nutrient-starved environment, promoting an increase in starch content, it was decided that adoption of this application would be challenging due to the large infrastructure required and the limited size of the designed pilot farms. Our interviews with the surrounding municipality representatives also revealed that it would not incentivize collaboration as effectively as animal feed or fertilizer, which have direct 71 agricultural benefits. It is recommended that further research on the production of biofuel be focused on low-cost methods of synthesis, as well altering farm designs to allow for greater biomass production. If an efficient method is found, there may be reason to utilize duckweed in this manner, as heavy metal concentrations within the wastewater may make other applications impractical. Further research should also be centered around improving the optimization of biomass production as well as the development of starch content within harvested duckweed. As of now, our current findings suggest that this application would not be economically beneficial to our stakeholders.

We also recommend looking further into the effect heavy metal concentrations have on the duckweed. This is an area of interest due to heavy metal contamination eliminating the viability of duckweed-based animal feed and fertilizer, requiring the biomass to be applied elsewhere. This is to avoid biomagnification within animals and crops that would be used as sources of food. Prior to harvest the duckweed should be observed for signs of chlorosis (Figure 13), in which the leaves of the duckweed become pale. The progression of chlorosis within the duckweed can reveal possible heavy metal contamination without the need for spectrophotometric analysis. However, despite chlorosis offering an easy way to estimate potential heavy metal contamination, it is recommended that prior to the first use of harvested duckweed and on a regular basis, the harvested duckweed is tested for heavy metal contamination using spectrophotometric methods to ensure that contaminated duckweed isn’t being used for animal feed or fertilizer. 72

Farm Design Recommendations

Along with the species of the duckweed and the post-harvest applications, we have recommendations for the design and location of each duckweed farm. The plot of land for the duckweed farm should be either inexpensive, or flat enough that the initial construction cost is minimal. Any plot of land at risk of flooding should not be considered. If the land is subject to flooding, the pond water may carry duckweed into the lake. This could result in duckweed entering the ecosystem of the lake. In order to prevent this, fences or nets may be installed around the pond so that the duckweed will not flow out of the pond. The farm could also be installed a farther distance from the river or lake to reduce this risk, but the cost of installation will increase, as the water will have to be diverted a farther distance.

The pond itself should follow some regulations. In order to efficiently harvest the duckweed, the pond should be narrow. We recommend a width between 4.5-6 m (15-20 ft). Anything wider than that will prove difficult and inefficient to harvest. Similarly, the depth of the pond should not be more than 5 ft deep (1.5 m). This will also ease the harvesting process, and should not inhibit the growth of the duckweed. For the harvesting process we recommend that pond rakes are used as automated methods are too large an investment for the communities where limited funds are available. Simple metal rakes may be used for harvesting duckweed instead of pond rakes if the prongs are close enough to pull duckweed into clumps that can be removed from the water. Once the duckweed has been harvested, we recommend placing it onto black plastic sheets in the sun with the duckweed evenly spread across it. The duckweed should remain under the sun on the plastic for several days, and in the event of rain it should be stored away from where it cannot get wet. The harvesting and drying processes could also be upgraded but would require sustainable investments and should only be considered once the duckweed farms have been proved to be productive and effective. If the funding is obtained to make upgrades, we recommend that the first upgrade is to automate the harvesting using the Proskimmer previously mentioned or a similar automated harvester. This would allow the farmer to extend the pond, and harvest with minimal labor 73 costs. In order to upgrade the drying process, we recommend using the oven drying technique. The ovens will be able to dry the duckweed within a day.

For stakeholders that have more than the minimum level of resources or for those that receive additional funds, several options exist to improve the pond. The first method of upgrading the pond would be to upgrade the water entry system. In order to do this, a pump system could be installed to transport the water at an adjustable rate through a series of pipes. This pump would have to be installed by a professional and the price will vary depending on the pond location and the pump quality. This pump will ensure that the flowrate will be controlled. Another upgrade would be adding a concrete floor to the pond, greatly reducing the erosion rate. This could only be done during the installation process, unless the pond were drained after installation. This would cost approximately $1600 for the materials, labor not included

(Concrete Cost Calculator, 2020). The last plausible upgrade would be to install a dam type apparatus to the exit canal. This would allow the farmer to plug the pond shut. This would allow the farmer to contain and control the pond to a higher extent. This also removes the need for a net or wire mesh/fencing. These upgrades will need to be able to withstand Amrenian winters and potential flooding. Due to this, we were not able to find a specific cost for these upgrades. 74

Conclusions

In this project we assessed the feasibility of a duckweed-based phytoremediation system in order to reduce nitrate and phosphate concentrations flowing into Lake Sevan from mouths of the Gavaraget,

Masrik, and Martuni rivers. By absorbing nutrients from the water before it flows into the lake, duckweed reduces the likelihood of eutrophic events from occurring. Once implemented, this system will also reduce wastewater runoff into rivers caused by agricultural and industrial development, preventing further damage to the surrounding ecosystem. From interviews conducted with local municipality representatives, it was also found that the algae blooms caused by eutrophication go beyond just ecological damage, presenting widespread social and economic consequences as well. To address these problems, not only were locations in need of treatment identified, but farms were designed to provide economic opportunities for locals who commit to their maintenance. Using the estimated biomass calculations and the research compiled regarding post-harvest applications of duckweed biomass, the water purification system is not only ecologically beneficial but also has the opportunity to be financially lucrative for the surrounding populace.

Future efforts should be focused on the further testing of the outlined pilot farm in order to corroborate the calculated biomass and growth rate estimations. Local concerns about the effectiveness of the system for water purification and estimated economic benefit can be relieved through small scale implementation and testing. Once completed, interested farmers and volunteers can be identified for support of larger scale farm development, and alterations can be made to the initial farm design in order to be more efficient and beneficial.

Successful and widespread implementation of duckweed farms in the tributaries in Lake Sevan will have profound environmental and economic impacts on Armenia. The removal of excess nutrients from the tributaries and reduction of wastewater runoff through duckweed farms will help reduce nutrient concentrations in Lake Sevan, preventing eutrophication. This will aid in the preservation of the delicate ecosystem of Lake Sevan Basin, allowing for businesses dependent on the lake to continue to prosper. In 75 addition, thousands of kilograms (several tons) of duckweed biomass could be produced allowing for cheap and sustainable sources of animal feed and fertilizer that offer several improvements over commercial animal feeds and fertilizers. These include increasing the growth rates of animals or reducing the runoff of nutrients from soil, further benefiting Armenians both economically and environmentally. If further investment is contributed to the project, stakeholders may also be able to produce large quantities of ethanol, an environmentally friendly and economically lucrative source of energy. 76

References

Acopian Center for the Environment. (2020, July 20). Acopian Center for the Environment.

https://ace.aua.am/

Aghajanyan, E. A., Avalyan, R. E., Simonyan, A. E., Atoyants, A. L., Gabrielyan, B. K., Aroutiounian, R.

M., & Khosrovyan, A. (2018). Clastogenecity evaluation of water of Lake Sevan (Armenia) using

Tradescantia micronucleus assay. Chemosphere (Oxford), 209, 1-6.

https://doi.org/10.1016/j.chemosphere.2018.05.142

Anthonius, C., Yong, A.S.K., & Fui, C.F. (2018). Supplementation of duckweed diet and citric acid on

growth performance, feed utilization, digestibility and phosphorus utilization of TGGG hybrid

grouper (Epinephelus fuscoguttatus x Epinephelus lanceolatus) juvenile. Wārasān Songkhlā

Nakharin, 40(5), 1009-1016. https://doi.org/10.14456/sjst-psu.2018.123

Appenroth, K.-J., Sree, K. S., Bog, M., Ecker, J., Seeliger, C., Böhm, V., Lorkowski, S., Sommer, K., Vetter,

W., Tolzin-Banasch, K., Kirmse, R., Leiterer, M., Dawczynski, C., Liebisch, G., & Jahreis, G.

(2018). Nutritional Value of the Duckweed Species of the Genus Wolffia (Lemnaceae) as Human

Food. Frontiers in chemistry, 6, 483-483. https://doi.org/10.3389/fchem.2018.00483

Aquamarine – aquatic weed harvesting equipment manufacturers for lake weed removal. (n.d.). Retrieved

February 26, 2021, from https://www.aquamarine.ca/

Babayan, A., Hakobyan, S., Jenderedjian, K., Muradyan, S., & Voskanov, M. (2006). (rep.). Lake Sevan:

experience and lessons learned brief (pp. 347–362).

Bandaru, N. (2020) [Green moss on brown soil] [Photograph]. Unsplash.

https://images.unsplash.com/photo-1594850285949-fb1b9a781150?ixid=MXwxMjA3fDB8MHx

waG90by1wYWdlfHx8fGVufDB8fHw%3D&ixlib=rb-1.2.1&auto=format&fit=crop&w=1068&

q=80 77

Blocks. T, & Rhoades, A. (2011). Thalloid and/or tiny floating plants. In Aquatic Plants of Pennsylvania:

A Complete Reference Guide (pp. 163-183). University of Pennsylvania Press, Inc.

Burt, T. P. (2019). Nitrogen cycle. In (Second Edition ed., pp. 135-142): Elsevier B.V.

Cheng, J., & Stomp, A.-M. (2009). Growing Duckweed to Recover Nutrients from Wastewaters and for

Production of Fuel Ethanol and Animal Feed. A. CLEAN-Soil, Water.

https://onlinelibrary-wiley-com.ezpxy-web-p-u01.wpi.edu/doi/abs/10.1002/clen.200800210

Concrete cost calculator | estimate average cost of concrete slab. (2020). Remodeling Calculator.

Retrieved May 5th, 2021 from https://www.remodelingcalculator.org/concrete-calculator/

Conley, D. J., Paerl, H. W., Howarth, R. W., Boesch, D. F., Seitzinger, S. P., Havens, K. E., Lancelot, C.,

& Likens, G. E. (2009). Ecology. Controlling eutrophication: nitrogen and phosphorus. Science

(American Association for the Advancement of Science), 323(5917), 1014-1015.

https://doi.org/10.1126/science.1167755

Cui W., Xu J., Cheng J.J., & Stomp A.M. (2011) Starch accumulation in duckweed for bioethanol

production. Biological Engineering, 3, 187– 197.

Cui, W., Cheng, J. J., & Appenroth, K. (2015). Growing duckweed for biofuel production: a review. Plant

biology (Stuttgart, Germany), 17(s1), 16-23. https://doi.org/10.1111/plb.12216 de Jonge, V., Elliott, M., & Orive, E. (2002). Causes, historical development, effects and future challenges

of a common environmental problem: eutrophication. Hydrobiologia(475/476), 1-19.

Dhote, S., & Dixit, S. (2009). Water quality improvement through macrophytes—a review. Environmental

monitoring and assessment, 152(1), 149-153. https://doi.org/10.1007/s10661-008-0303-9 78

Dredging 101. (n.d.). Retrieved March 17, 2021, from

https://www.dredgingspecialists.com/Dredging101.htm#:~:text=Hydraulic%20Dredging%20Cost

s.,average%20of%20the%20above%20costs

Engwa, G. A., Ferdinand, P. U., Nwalo, F. N., & Unachukwa, M. N. (2019). Mechanism and Health

Effects of Heavy Metal Toxicity in Humans https://doi.org/10.5772/intechopen.82511

Faizal, A., Sembada, A. A., & Priharto, N. (2021). Production of bioethanol from four species of

duckweeds (Landoltia punctata, Lemna aequinoctialis, Spirodela polyrrhiza, and Wolffia arrhiza)

through optimization of saccharification process and fermentation with Saccharomyces

cerevisiae. Saudi journal of biological sciences, 28(1), 294-301.

Frédéric, M., Samir, L., Louise, M., & Abdelkrim, A. (2006). Comprehensive modeling of mat density

effect on duckweed (Lemna minor) growth under controlled eutrophication. Water research,

40(15), 2901–2910. https://doi.org/10.1016/j.watres.2006.05.026

Hassan, M. S., & Edwards, P. (1992). Evaluation of duckweed (Lemna perpusilla and Spirodela

polyrrhiza) as feed for Nile tilapia (Oreochromis niloticus). Aquaculture, 2021(104), 11.

https//doi.org/10.1016/0044-8486(92)90213-5

Haustein, A. T., Gilman, R. H., Skillicorn, P. W., Guevara, V., Díaz, F., Vergara, V., Gastañaduy, A., &

Gilman, J. B. (1992). Compensatory growth in broiler chicks fed on Lemna gibba. British journal

of nutrition, 68(2), 329-335. https://doi.org/10.1079/BJN19920092

Haustein, A. T., Gilman, R. H., Skillicorn, P. W., Vergara, V., Guevara, V., & GastaÑAduy, A. (1990).

Duckweed, A Useful Strategy for Feeding Chickens: Performance of Layers Fed with

Sewage-Grown Lemnacea Species. Poultry science, 69(11), 1835-1844.

https://doi.org/10.3382/ps.0691835 79

Hegazy, A. K., Kabiel, H. F., & Fawzy, M. (2009). Duckweed as heavy metal accumulator and pollution

indicator in industrial wastewater ponds. Desalination and Water Treatment, 12, 7.

How much does it cost for land excavation and grading? (2021). Home Advisor Retrieved May 5, 2021

from

https://www.homeadvisor.com/cost/landscape/excavate-land/#:~:text=On%20average%2C%20yo

u'll%20end,upwards%20of%20%246%2C000%20to%20%2410%2C000.

Gupta, C., & Prakash, D. (2013). Duckweed: an effective tool for phyto-remediation. Toxicological and

environmental chemistry, 95(8), 1256-1266. https://doi.org/10.1080/02772248.2013.879309

Kabir, A. N. M. A., Hossain, M. A., & Rahman, M. S. (2009). Use of Duckweed as Feed for Fishes in

Polyculture. Journal of Agriculture & Rural Development J Agric Rural Dev J Agric Rural Dev,

7(1), 4.

Körner, S., Vermaat, J. E., & Veenstra, S. (2003). The Capacity of Duckweed to Treat Wastewater.

Journal of Environmental Quality, 32(5), 1583-1590.

https://doi.org/https://doi.org/10.2134/jeq2003.1583

Landolt, E. (1986). Biosystematic investigations in the family of duckweeds (Lemnaceae). II: The family

of Lemnaceae: a monographic study. 1. Veröffentlichungen des Geobotanischen Institutes der

ETH, Stiftung Rübel, Zürich(71).

Laplante, B., Meisner, C., & Wang, H. (2005). Environment as cultural heritage: The Armenian

Diaspora’s Willingness to Pay to Protect Armenia’s Lake Sevan. In (pp. 1-35). Washington, DC,

United States of America: World Bank. 80

Lasfar, S., Monette, F., Millette, L., & Azzouz, A. (2007). Intrinsic growth rate: a new approach to

evaluate the effects of temperature, photoperiod and phosphorus-nitrogen concentrations on

duckweed growth under controlled eutrophication. Water research, 41(11), 2333–2340.

https://doi.org/10.1016/j.watres.2007.01.059

Laube, H. R., & Wohler, J. R. (1973). Studies on the decomposition of a duckweed (lemnaceae)

community. Bulletin of the Torrey Botanical Club, 100(4), 238-240.

https://doi.org/10.2307/2484804

Law on the restoration of the Sevan Lake Ecosystem, No. ZR-190 (2001).

Leng, R. A. (1999). DUCKWEED: A tiny aquatic plant with enormous potential for agriculture and

environment. http://www.fao.org/ag/AGAinfo/resources/documents/DW/Dw2.htm

Leng, R. A., Stambolie, J. H., & Bell, R. (1995). Duckweed - a potential high-protein feed resource for

domestic animals and fish. Livestock Research for Rural Development., 7, Article 5.

Liu, Y., Xu, H., Yu, C., & Zhou, G. (2021). Multifaceted roles of duckweed in aquatic phytoremediation

and bioproducts synthesis. Global change biology. Bioenergy, 13(1), 70-82.

https://doi.org/10.1111/gcbb.12747

Li, Y., Zhang, F., Daroch, M., & Tang, J. (2016). Positive effects of duckweed polycultures on starch and

protein accumulation. Bioscience reports, 36(5). https://doi.org/10.1042/BSR20160158

Lind, D., & Taslakyan, L. (2005). Restoring The fallen blue sky: management issues and environmental

legislation for Lake Sevan, Armenia. Environs Environmental Law and Policy Journal, 31–58.

Martinez, S., Sáenz, M. E., Alberdi, J. L., & Di Marzio, W. D. (2020). Comparative ecotoxicity of single

and binary mixtures exposures of cadmium and zinc on growth and biomarkers of Lemna gibba.

Ecotoxicology (London), 29(5), 571-583. https://doi.org/10.1007/s10646-020-02213-4 81

McCrackin, M., Jones, H. P., Jones, P. C., & Moreno-Mateos, D. (2017). Recovery of lakes and coastal

marine ecosystems from eutrophication: A global meta-analysis. Limnology and oceanography,

62(2), 507-518. https://doi.org/https://doi.org/10.1002/lno.10441l

McLay, C.L. (1976), The effect of pH on the population growth of three species of duckweed: Spirodela

oligorrhiza, Lemna minor and Wolffia arrhiza. Freshwater Biology, 6: 125-136.

https://doi.org/10.1111/j.1365-2427.1976.tb01596.x

Meyers, L. (1982). Composition, Biomass, and Protein Content of a Lemna trisulca L., Invertebrate

Community in a Prairie Wetland South Dakota State University.

http://openprairie.sdstate.edu/etd/183

Miranda, A. F., Biswas, B., Ramkumar, N., Singh, R., Kumar, J., James, A., Roddick, F., Lal, B., Subudhi,

S., Bhaskar, T., & Mouradov, A. (2016). Aquatic plant Azolla as the universal feedstock for

biofuel production. Biotechnology for biofuels, 9(1), 221.

https://doi.org/10.1186/s13068-016-0628-5

National Statistical Service of RA. (2013). The results of 2011 population census of the republic of

armenia (Indicators of the republic of armenia) / statistical committee of the republic of armenia.

Statistical Committee of the Republic of Armenia. https://www.armstat.am/en/?nid=53

Nazaryan, L. (2007). Armenian Dried Fruit is Not Competitive. hetq. Retrieved March, 13 from

https://hetq.am/en/article/24027

Onuki, S., Koziel, Jacek A., van Leeuwen, J., Jenks, William S., Grewell, David A., & Cai, L. (2008).

Ethanol production, purification, and analysis techniques: a review. Agricultural and Biosystems

Engineering Conference Proceedings and Presentations. 68. 82

Pal, D., & Maiti, S. K. (2019). Evaluation of potential human health risks from toxic metals via

consumption of cultured fish species labeo rohita: a case study from an urban aquaculture pond.

Exposure and health, 11(1), 33-46. https://doi.org/10.1007/s12403-017-0264-8

Papadopoulos, F., Metaxa, E., Latrou, M., & Papadopoulos, A. (2014). Evaluation of performance of

full-scale duckweed and algal ponds receiving septage. Water Environment Research, 86(12),

2309-2316.https://doi.org/10.2175/106143014X14062131178754

Peeters, E., Van Zuidam, J., Van Zuidam, B., Van Nes, E., Kosten, S., Heuts, P., Scheffer, M. (2013).

Changing weather conditions and floating plants in temperate drainage ditches. Journal of

Applied Ecology, 50(3), 585-593

Proskim duckweed and watermeal removal. (n.d.). Retrieved February 26, 2021, from

https://www.proskim.com/

Pulido, C. R. F. (2016). DUCKWEED AS A SUSTAINABLE SOIL AMENDMENT TO SUPPORT CROP

GROWTH, ENHANCE SOIL QUALITY, AND REDUCE AGRICULTURAL RUNOFF [The Pennsylvania

State University].

Rooijakkers, P. (2016). Photosynthesis model to predict duckweed growth at the Ecoferm greenhouse.

Wageningenur.

Tanner, C. C., & Headley, T. R. (2011). Components of floating emergent macrophyte treatment wetlands

influencing removal of stormwater pollutants. Ecological engineering, 37(3), 474-486.

https://doi.org/10.1016/j.ecoleng.2010.12.012

The Pond Guy. Pond and water garden supplies. (n.d.). Retrieved February 26, 2021, from

https://www.thepondguy.com/ 83

Thompson, C. L. (2010). Cellulosic Ethanol: the Benefits, Obstacles, and Implications for Nebraska. van den Berg, H., van Lidth de Jeude, E., Hak, A., Mooren, C., Leen den Uigl, R., Giomi, T., &

Kraaijeveld, S. (2015). Duckweed, a tiny aquatic plant with growing potential: A research on the

potential applications of duckweed in urban water systems in The Netherlands.

https://www.hhdelfland.nl/overheid/onderhoud-aan-water/Duckweedatinyaquaticplantwithgrowin

gpotentialStudentenUniversiteitvanUtrecht2015.pdf van den Top, N. 2014. Dynamic modelling of duckweed production on the first ECOFERM! closed-cycle farm for rosé calves.

Vinçon-Leite, B., & Casenave, C. (2019). Modelling eutrophication in lake ecosystems: A review.

Science of The Total Environment, 651, 2985-3001.

https://doi.org/https://doi.org/10.1016/j.scitotenv.2018.09.320

Wentzel, K., & Gharabegian, A. (June 19, 2009) A bio-ethanol fuel program for Armenia, Asbarez.

Retrieved from https://asbarez.com

Wills, M., & Irvine, K. N. (1996). Application of the National Sanitation Foundation Water Quality

Index in the Cazenovia Creek, NY, Pilot Watershed Management Project (No. 95–104). Middle

State Geographer.

Zhao, Z., Shi, H., Liu, Y., Zhao, H., Su, H., Wang, M., & Zhao, Y. (2014). The influence of duckweed

species diversity on biomass productivity and nutrient removal efficiency in swine wastewater.

Bioresource technology, 167, 383-389. https://doi.org/10.1016/j.biortech.2014.06.031 84

Appendix A - International Reports of the Ministry of Nature Protection with Wetland Measurements (Babayan et al., 2006)

Reports to International Treaties Measures Projected (P) and Implemented (I) in the Wetlands

First National Communications of the Republic of Recognition of the role of Lake Sevan in climate Armenia under the UNFCCC (First National mitigation and greenhouse gas effect reduction Report, 1998) without special measures proposed.

Armenia National Environmental Action Program (P) Development of integrated water resources (Main Report, 1999) master plan. (I) Undertake pilot project for restoration of Lake Gilli (Successful).

Biodiversity of Armenia (First National Report, Special attention is given to conservation of 1999) wetland biodiversity and wetland landscapes, in particular in Lake Sevan Basin, without special measures particularly related to the wetlands.

Biodiversity Strategy and Action Plan of Armenia (P) Provide technical assistance and equipment to (Main Report, 1999) Sevan National Park. (P) Establish new protected areas. (I) Conserve and rehabilitate key wetland ecosystems, including the ecosystem of Lake Sevan (Unsuccessful). (P) Develop individual action plan for conservation of endangered fish ishkhan. (I) Promote sustainable fisheries (Unsuccessful). (I) Wetland management training (Successful).

Lake Sevan Action Program (Main Report, 1999) (I) Undertake pilot project for restoration of Lake Gilli (Successful). (I) Rehabilitate ishkhan hatcheries (Unsuccessful). (I) Develop a crayfish fishery (Successful). (I)Collect environmental data (Successful).

Natural Resources Management and Poverty Gegharkunik Marz is one of two case areas of Reduction Project (Ongoing) project implementation. In particular, the management plan for Sevan National Park should be prepared and implemented. 85

Appendix B - Guiding Questions for Interviews with Municipality Representatives

Introduction

The American University of Armenia’s Acopian Center for the Environment is sponsoring a project that seeks to implement a duckweed-based water treatment system within the tributaries surrounding Lake

Sevan. The goal of this system is to reduce the level of nutrients feeding into Lake Sevan, preventing further eutrophication events and preserving the lake’s ecosystem. The project will also seek to benefit the municipalities surrounding the implemented duckweed farms through the various post-harvest applications of duckweed biomass, such as animal feed or fertilizer. Due to the project needing the continued support of the local communities in order to have this desired impact, we would like to ask you a few questions in order to gauge your interest in this endeavour.

1. Would members of your municipality be willing to volunteer on a project working to prevent

further ecological damage from occurring in Lake Sevan?

2. Would members of your community be willing to support a water treatment project if it led to the

development of one of these post-harvest applications of duckweed farming?

3. Are there any logistical challenges that may arise in the implementation or management of these

farms?

4. Is your municipality able to offer any resources towards the building of the farms (i.e backhoes)

or maintenance of the duckweed farms?

5. Are there any zoning laws or other legal challenges that you are aware of that might impact the

building and implementation of a duckweed farm?

6. Are there any riverside locations that have flat/cheap land where an artificial pond could be

installed? 86

Appendix C - Algorithm Variable Information (Rooijakkers, 2016)

Name Value Unit Definition

Cn Input mgP/L N concentration

Cp Input mgP/L P concentration

D Calculated g/m²s mat density at time t

D0 100 g(dry)/m²s initial mat density

DL 176 g(dry)/m²s limit mat density

E Input h photoperiod

Emin 2 h minimal photoperiod

Eop 13 h optimal photoperiod g Calculated 1/d growth rate for pH gmax 0.27 1/d maximum intrinsic growth rate for pH

Kin 604 mgN/L N inhibition rate

Kip 101 mgP/L P inhibition rate

Kn 0.95 mgN/L N saturation rate

Kp 0.31 mgP/L P saturation rate p Input - pH

R 0.62 1/d maximum intrinsic growth rate constant ri Calculated 1/d intrinsic growth rate ri(E) Calculated 1/d growth rate for photoperiod ri(P,N) Calculated 1/d growth rate for P and N ri(pH) Calculated 1/d growth rate for pH ri(T) Calculated 1/d growth rate for temperature rimax 0.45 1/d maximum intrinsic growth rate rs Calculated 1/d specific growth rate

T Input ℃ temperature t Input s retention time

Top 26 ℃ optimal temperature

θ1 0.66 - non-dimensional constant smaller than 1

θ2 0.0025 - non-dimensional constant smaller than 1

θ3 0.0073 - non-dimensional constant smaller than 1

θ4 0.65 - non-dimensional constant smaller than 1

훼E 0.42 1/d constant for the maximum intrinsic growth rate

훼P,N 0.46 1/d constant for the maximum intrinsic growth rate

훼T 0.41 1/d constant for the maximum intrinsic growth rate 87

Appendix D - Ethanol Production Calculations

Ethanol per gram of dry Estimated Biomass (kg) duckweed (g) Density of Ethanol (g/L) Lower Limit 754.78 0.16 789.3 Estimated Output [(754.78) (.16) (1000)] / 789.3 = 153.002 L Upper Limit 1093.43 0.258 789.3 Estimated Output [(1093.43) (.258) (1000)] / 789.3 = 357.412 L