Tuesday, August 16, 2016

A Tale of Two Watersheds

Interdisciplinarity was the nature of my work this summer, looking to see if there was a correlation between social distance and geographic distance within the farmer social network. The farmer social network is a visual diagram that was created by the sociologists from last year’s LAKES team that shows which farmers talk to each other about farming practices within the Red Cedar watershed. Small dots called nodes represent people, while lines connecting them called edges represent their connection.
 Surveys that were sent to farmers create the social network. These surveys ask about best management practices (BMP), as well as interest in learning about conservation agriculture, and who the farmers talks to about their land use practices.
Why does this information matter? By seeing the lines of communication between farmers, we can see who would be open to learning about BMPs for their farms, and how to effectively disseminate information about conservation agriculture.

While the sociologists were able to create a fantastic social network, they did not look at it in terms of geography. That is, seeing if there is a correlation between social distance and geographic distance. I started my work by getting familiar with the survey that created the farmer social network, as well as prepping the current data for statistical analysis in a geographic information system (GIS) software called ArcMap. I also recreated the farmer social network in terms of geography.
In AcrMap, I looked to see if there was a spatial autocorrelation between any of the questions in the survey. Spatial autocorrelation measures the clustering of values within a map.
As for the new social network, I made the sub-watersheds within the Red Cedar into nodes from which to anchor the social network. Edges were made, connecting to the farmers within their boundaries. By viewing the social network as such, I could see if there was any clustering of social connection based on sub-watershed.

From my analysis of the farmer social network, I found that there is a contrasting relationship between two sub-watersheds within the Red Cedar.
On one hand, the Lower Pine Creek-Red Cedar shows a high clustering of farmers who would be interested in learning about conservation agriculture, soil health, and economic projections for their farms, while at the same time, having a high clustering of BMP usage.
 On the other hand, Hay River showed a randomization in those interested in educational programs, as well as randomization in BMP usage in general. 

Where do we go from here? My findings open the door for many different avenues of research. From what I’ve found, we can start asking: what are the differences between these two sub-watersheds in terms of geography, and society? What is working for one, and why is it not working for the other?

            With this information, we can better learn the factors that affect the farmer social network in the Red Cedar, and how to better disseminate information about conservation agriculture.

Friday, August 12, 2016

The Logic of the Lakes: Economic Research on Preferences and Perceptions

When I began my research with the economics team, I assumed that we would find answers that fit neatly into the box of questions that I had formulated. I expected to find particular trends and significant factors, and in the back of my mind this project was a way to prove my preconceived notions true. Eight weeks later, and I have more questions than answers. But, as they say, the best research often ends this way.

Personally, I was drawn to the willingness to pay questions we designed on our survey. People's perceptions and behavior fascinate me because often times they are not as rational as assumed. There is always a story behind the decisions people make, especially when it comes to environmental goods, as there is no explicit price to monitor and construct the story for us. So we need dig deeper to fill in the gaps ourselves.

Wednesday, August 10, 2016

Oral History of Lakes Menomin and Tainter

This summer, I focused on oral history of the Red Cedar Watershed, and more specifically, Lakes Menomin and Tainter. But what is oral history, you ask? Oral history is much like written history; it has its own slants and perspectives depending on the source. Unlike written history (you might have guessed), it is spoken. In this way, one can gather an understanding of the past as well as the significance that people place on certain events depending on how much or little they talk about them. I used this concept - the way people talk about the past affects their actions in the present - to understand how people perceive the lakes and how they have changed.

A postcard from the mid 1900s featuring "the old swimming hole" at Wakanda Park.
That means that over a couple of months, I interviewed more than 10 people in the Red Cedar Watershed, read interview transcripts conducted by previous REU students, attended various public meetings relating to the lakes, did archival research, and took a lot of field notes. As I and my research partner, Madison, started wrapping up data collection, we also started analyzing the data we gathered. The data filled various roles in my research. I analyzed the interviews to get an idea of how people talk about the lake health as they’ve known it. I used the information I found in the archives to fill in the spaces of what people didn’t say in interviews. That is, people tell things as they know it. The archives proved to be a useful resource that allowed me to gather a fuller representation of the lakes’ history. By participating in public meetings, I experienced what sort of things are being done in the present to address lake health.

After all of that - data gathering, coding, and analyzing - I drew a few conclusions.
  1. Personal experience, strength of memories, and sense of place influence how they perceive current lake health, what they think should or can be done about declining lake health, and how they are involved in lake clean up efforts.
  2. Based on the interviews conducted, the dominant theme is that the lake has gotten worse. Most people expressed that they were unsure of how to move forward in addressing declining lake health, even if they are involved in a lake protection organization.
  3. Even though uncertainty exists for a variety of stakeholders, steps are being taken to slow the declining lake health.

So, people that have lived in the area certainly had stories to share, but what comes with that is a sense that the lakes have gotten worse, and there’s too much to do in order to reverse that decline, and that it’s an insurmountable problem. Result: little movement forward.
There was also a sense that the lake has always been green, and we don’t know how to begin to address that, and that it’s an insurmountable problem. Result: also little movement forward.
There was even a small sense for some that the lakes have improved, and there isn't anything to be done. Result: you already know - little movement forward.

Why does this matter? So what - some community members feel frustrated about lake health and don't know how to proceed. Can't we move ahead anyway? Well, yes, you can try to do that. But a very important part of this is that the lakes are a part of the community, and it's unlikely that people in the community will choose to be involved with something that feels like an uphill battle. This presents a disadvantage to all parties. Instead, make it easy for the community to understand what is going on. Work to build a consistent community understanding about the lakes' present health. Host educational events about the history of the lakes in order to establish a consistent, unified narrative in the community about the lakes' history as a community center and a clear vision for the future. In this way, people can better comprehend the situation, and are more likely to come together for such a worthy cause.

Socializing for a Cleaner Watershed: Social Connectivity and BMP Lease Agreements

It’s been said that we are the product of the five people with whom we spend the most time.  In my case, I am (very luckily!) the composite of my kind and talented research partner, Clare Salerno, and my brilliant and enthusiastic research mentor, Dr. Nels Paulson, as well as my two charming roommates and, of course, the houseplant that sat atop my desk as I researched Non-Operating Landowners and BMP Lease Agreements in Dunn and Baron County, WI at the University of Wisconsin-Stout this summer.

While the idea of reflecting the values and characteristics of the people we hang around with most seems obvious, it has powerful implications in every scenario to which it is applied—be it the workplace, personal relationships, or otherwise.  In the case of our research, we found that Non-Operating Landowners who have close relationships with their tenants and are connected to other Non-Operating Landowners through groups and organizations (such as Farmer’s Unions, sportsman’s clubs, and church groups) place a higher value on the preservation of their farmland and are more likely to include Best Management Practices in their lease agreements with the farmers who rent out and farm their land.

Monday, August 8, 2016

Identifying Changes in Land Use and Highly Erodible Soil

As part of the geography team, my task was to map highly erodible soils, and to get an estimate of how much phosphorus went into the waterways due to land use change from 2011 to 2015. This was done using a Geographic Information System (ArcGIS), which is computer software used to map and analyze spatial data. GIS data comes in the form of layers, which contain an attribute table describing the features of the data.

To map highly erodible soils, I first had to download the soils data layer for the state of Wisconsin. This layer uses polygons to visualize the shape and area of the different types of soils. It contained a field in its attribute table with a code for its soil type. These codes vary by county, so in order to find out which ones were highly erodible, I contacted NRCS district conservationists for counties within the Red Cedar watershed and asked them for their list of codes for highly erodible soils. I then selected the different codes from the table using a tool called Select by Attribute, and exported that data into a new layer containing only highly erodible soils. I found there were about 418,000 acres of highly erodible soils in the Red Cedar watershed.

To estimate how much phosphorus was going into the waterways due to changes in land use, I had to find out how much land changed from grassland to cropland. The reason behind this was that no topsoil is lost when there is grass or pasture, but when the land gets farmed and plowed, there is significant loss of topsoil. Plowing leaves the soil exposed to the elements and when it rains, this soil erodes and washed out into the waterways. This soil contains phosphorus, which is what causes the algal blooms. To calculate the change between these two types of land cover, I downloaded satellite imagery from 2011 and 2015 from the USDA. This imagery was classified according to the type of land cover and the different types of crops. I reclassified the two images into two classes, grassland and cropland. I then input these two images in Image Analysis and used the Difference tool, which resulted in a map with three classes, change from grassland to cropland, change from cropland to grassland, and no change. I then calculated the area and found that 8,738 acres changed from cropland to grassland, and 82,047 acres changed from grassland to cropland. Assuming farming at state standards, this means 1,312,752 tons of topsoil were lost from 2011 to 2015. Taking the average soil P test for Dunn and Barron counties of 52ppm, we get 75.25 tons of added phosphorus to waterways. Finally, I reclassified one more time to get only cropland, and overlayed this layer with the highly erodible soils layer and calculated the area. I found that about 85,000 acres, (20%) of highly erodible land are being farmed on.

The maps I produced can help illustrate what happens when there is a big increase in farming. This increase can potentially be explained by rising corn prices. My research found that enrollment in conservation programs such as CREP decreased as prices in corn increased. My map of farmland on highly erodible soil can also help in identifying which areas to target for a bigger push in conservation agriculture.


Mathematical Modeling to Forecast Chlorophyll Levels and Algae Blooms

I worked on mathematical models for forecasting chlorophyll concentration and finding a condition for blooming. Mathematical models help with understanding important processes that govern chlorophyll concentration.  These models are analyzed by considering all relevant variables on the same time and length scale to determine the relative impact of each physical process. This normalization yields non-dimensional parameters that assist in the interpretation of analytic and computational results. We analyzed data from past REU students and the DNR to use in our models.We modeled chlorophyll growth because there is a direct relationship between chlorophyll and algae counts because algae produces chlorophyll as a by-product of photosynthesis and chlorophyll is more easily measured.
The forecasting model predicts how long after a large flushing event it takes for the lake to turn green and smelly if there’s no rain. We used a logistic growth model to capture longer term algae growth and a carrying capacity was estimated by looking at how much the algae can grow if there’s no nutrients for them to consume. We ran 100 simulations varying the carrying capacity and then the growth rate. We found the most variation when the carrying capacity was varied but in all simulations we found that it takes about 3 weeks to hit the carrying capacity level, which corresponds well with data collected from past REUs and the DNR.
The bloom condition was found by solving an equation that describes chlorophyll growth over time based on its growth and movement. The bloom condition was found by solving the equation analytically and was found to rely on Secchi depth and the ratio between growth rate and turbulence. The equation was also solved numerically using computer programming and found very similar results to the analytic solution.  

 Both of these models can take proposed solutions to the algae problem and test how effective they would be at preventing blooms. Both models advocate the solution of increasing turbulence and flow in the lake and the bloom condition models also suggests lowering the algae growth by limiting phosphorus and nitrogen to help solve this problem. 


Intersecting Identities and Land Meanings
BMP Use Among Non-Operating Landowners

Forty percent of farmland in the United States is rented out, mostly from Non-Operating Landowners (NOLs), or people that own land but do not themselves farm it. Nearly ⅔ of these NOLs are over the age of 65.  NOLs represent a sizeable, but vastly understudied, portion of landowners in the United States, and my research as part of the Sociology team this summer focused on use of conservation agriculture Best Management Practices (BMPs), such as conservation tillage or buffer strips, on farmland owned by NOLs.  
Our initial research questions sought to discern how factors like age, gender, conservation values, and relationship with one’s tenant affect BMP use and BMP lease agreements.  These questions guided our decision to talk to widowed women NOLs as well as framed the questions we asked NOLs in our survey.  Ultimately, my project aimed to discern how the intersecting identities and meanings NOLs associated with their land, as collected through survey and interview data, predict the levels of BMP use on their land.