Topic of Interest

From the day I figured out how to walk and use my hands, I had my life jacket on my chest and a tacklebox by my side. I would fish every single day, rain or shine if it meant I could be outside. It didn’t matter the trek to get there. If I had to hike through a forest or walk across my yard, I would figure out a way to be on the water. Fishing started as a way to be outside with my dad, my grandpa, and my friends. But, that hobby turned into a form of adventure that kept me wanting more and more. Everyday I would learn something or experience something new! My lust for adventure turned into a full blown addiction that would never slow down. I did research for 5 hours a day and practiced my casting in my basement. Subscriptions to magazines and youtube channels quickly piled up. And at this point I got very interested in the practice of fish stocking done by the Department of Natural Resources.

There are more lakes in Wisconsin than than any other state aside from ALaska which is nearly 10X the size of Wisconsin. There are more than 15,000 lakes in Wisconsin and is home to some of the best fishing in the world! Don’t take my word for it, do some research for yourself and you will find some world record setting fish. But, these fisheries are only as good as they are because they have been maintained and are well regulated. The Department of Natural Resources (DNR) stocks fish into the lakes every single year to ensure that fish populations and sizes remain in healthy condition. Fish stocking is essentially taking very young fish species and dumping them into the lake so that they can breed and grow a natural population of fish.

Data

I am retrieving data from the Wiscosin Department of Natural Resources Fish Stocking Summaary page which provides a table data on each fish stocking event. I specifically will be looking into the County of Waukesha because that is where I am from and want to get a better understanding of my local lakes and ponds.

Information dates all the way back to 1972 when the fish stocking efforts became standardized by the DNR and tracked. The data will be exctracted from the website below:

https://cida.usgs.gov/wdnr/apex/f?p=220:1:::NO::P1_COUNTY_NAME:WAUKESHA&cs=1053AA4B147E872192D7129B1A3EAB620

Then the data will be uploaded as a csv file for further analysis an to answer my research question, prove my hypothesis, or disprove my hypothesis.

Research Question and Hypothesis

Question: The question I want to determine is: Which lake in Wisconsin received the most amount of fish through the DNR’s fish stocking efforts in 2020? Which fish were stocked the most in Waukesha County?

Hypothesis: I am predicting that the lake lakes that receive the most amount of fish will tend to be the bigger lakes like Pewuakee Lake, Oconomowoc Lake, and Lac La Belle. I also predict that the most common fish the DNR is stocking in the lakes are Walleye and Muskies since they are the two most commonly sought after sportfish in the Midwest. Muskies are known for their unrelenting force, elusiveness, incredible stature. Walleyes are a delecacy in the fishing world with their tastey white flakey flesh. They also are known to grow very big and have a feisty nature.

Loading the Data

Data Dictionary

Here is a list of the column headers and what they mean below:

Year - the year the stocking information occured during

StockedWaterbodyName - the name of the lake the stocking occred in

LocalWaterbodyName - the nickname the town or area gives the lake

Location - the coordinate location of the stocking

Species - the species of fish, i.e. Walleye, Musky, Catfish, etc.

Strain..Stock. - what other body of water or where the fish came from to be stocked

Age.Class - how old the fish is determined to be in a classification form

NumberFishStocked - total amount of fish stocked in lake that year of that species

Avg.FishLength..IN. - average length of the type of fish stocked in that lake by year

Packages for Analysis

The packages I will need in my analysis are as follows:

Dplyr - allows me to create code chunks necessary for analysis

tidyverse - allows me to use many functions necessary for creating visualizations

xml/httr - allows me to perform data scraping in secondary data for further analysis

magrittr/Rcurl/rvest - allows for further analysis

Visualization and Analysis

The Bar Chart above shows a visualization of the total amount stocked in Waukesha Wisconsin, broken down by the fish species. This data has not been limited by data range at all and includes all stocking years since 1972. I would find this as useful data because you can see which fish have needed the most care by the DNR and in some ways which fish are having the hardest time reproducing on their own.

Some fish have trouble reproducing on their own because they have more natural predators or fishing pressure, but, this can also point to environmental hazards that we didn’t even know were there. For example, trout species are very sensitive to water conditions. Any water that is too basic or acidic can affect the longevity of live for trout species. It can also point to areas that have lower dissolved oxygen levels do to farming, fertilizing, or dying vegetation. This helps DNR representatives determine where they need to focus their time, energy, and costs.

## # A tibble: 21 x 4
## # Groups:   Year, Species [5]
##     Year Species       `Fish Stocked` `Lake Name`        
##    <int> <chr>         <chr>          <chr>              
##  1  2020 NORTHERN PIKE 82             WILLOW SPRINGS LAKE
##  2  2020 BROOK TROUT   8,000          MUKWONAGO RIVER    
##  3  2020 WALLEYE       7,046          LITTLE MUSKEGO LAKE
##  4  2020 RAINBOW TROUT 6,011          NAGAWICKA LAKE     
##  5  2020 BROOK TROUT   6,000          SCUPPERNONG RIVER  
##  6  2020 RAINBOW TROUT 5,934          LOWER GENESEE LAKE 
##  7  2020 WALLEYE       5,537          UPPER NEMAHBIN LAKE
##  8  2020 RAINBOW TROUT 5,016          UPPER NEMAHBIN LAKE
##  9  2020 WALLEYE       483            FOWLER LAKE        
## 10  2020 BROWN TROUT   4,959          LOWER GENESEE LAKE 
## # ... with 11 more rows

The table above provides a good visualization of the number of fish stocked in each lake in Waukesha County (Year 2020). I want to use this information in order to find where the most fish get stocked. This will help my chances next time I go fishing. Obviously the more fish that are in each lake then the better probability that you will catch more fish. The table also breaks out the fish species that get stocked in each lake. This is important because it helps if I am looking to fish for a specific kind. I don’t want to limit this data table by taking out certain species of fish, because the species I fish for depends on the day and how I am feeling.

From this table summary, I determined that Pewaukee Lake and Okauchee Lake had the most walleye stocked in their lakes in the year of 2020. I am trying to do more Walleye fishing this year and will target those lakes a lot more with this information. Brook Trout or Brookies are also extremely popular fish to target among the fly fishermen in Wisconsin. I can see from the table that most brook trout get stocked in rivers and creeks. More than 14,000 brook trout were stocked in 2020, which is great for the fishermen of Wisconsin.

This table is interesting because we can determine from this information which fish typically survives the longest - based on the age they are released by the DNR.

One of the age classes that gets released the least is the adult fish. This makes sense to me because adult fish that have been raised their entire lives in a fish tank will not survive in the wild. The fish class that is released most often is Yearlings which only survive a small portion of their lives in fish tanks. Furthermore, yearlings have the size and maturity to protect themselves from predators when they are stocked in lakes.

This graph shows a depiction of where the fish come from before they were stocked into new lakes. We can see that one of the lakes/rivers they came from was the Erwin Chain of Lakes in Wisconsin. This is good information for anglers and DNR representatives because it tells us that that cahin of lakes already has a very healthy population of fish. That chain of lakes is already populated enough and healthy enough that it does not need a significant amount of stocking to remain healthy. In fact, the chain of lakes is healthy enough that the DNR can remove a large amount of fish from those bodies of water.

## # A tibble: 23 x 3
## # Groups:   Species [5]
##    Species                     Avg.FishLength..IN. Age.Class       
##    <chr>                                     <dbl> <chr>           
##  1 MUSKELLUNGE                                15   YEARLING        
##  2 RAINBOW TROUT                              15   ADULT           
##  3 NORTHERN PIKE                              14.1 LARGE FINGERLING
##  4 MUSKELLUNGE                                13.3 FINGERLING      
##  5 NORTHERN PIKE                              12.8 LARGE FINGERLING
##  6 MUSKELLUNGE                                12.4 LARGE FINGERLING
##  7 RAINBOW TROUT                              11.6 YEARLING        
##  8 RAINBOW TROUT                              11.6 YEARLING        
##  9 NORTHERN PIKE X MUSKELLUNGE                11   FINGERLING      
## 10 RAINBOW TROUT                              11   YEARLING        
## # ... with 13 more rows

This is an insightful table that tells us a bit about what type of fish get released during their different stages of life. I filtered the data to only display fish that were released over 10 inches. We can use this table to add to our 3rd visualization breaking down age class.

Muskies can grow to be taller than most middle school kids and can weigh about as much as a German Shepherd dog. From this table we can see that the largest average size muskies get stocked at is 15 inches long. We can make the inference that Muskies most likely get released as yearlings or fingerlings. This makes sense as our bar chart clearly showed yearling and fingerling age classes are the most likely to get stocked.

At one point in time, we saw that Rainbow Trout were recorded to be released on an average of 15 inches. What we know about Rainbow Trout is that they do not grow much bigger than 25 inches during a typical lifetime. We can see that they are released most often as yearling fish.

I find this really fascinating because Muskies are able to survive as younger fish when they are stocked when compared to other fish like Rainbow Trout and Brook Trout.

Secondary Data For Further Analysis

I gathered data from Twitter’s API on Wisconsin DNR and their WDNR twitter handle in order to find the sentiment regarding the organization. As a fisherman, DNR are understood to be the “fun stoppers” on the water. They will drive up to you on the water, incspect your boat, check for boaters and fishing liscense, and can sometimes spoil a great fishing spot.

Our primary data suggests they are doing so much to protect the species we love to fish for like Walleye, Muskies, and Trout. But if we run a sentiment analysis we can see how appreciative the folks in Wisconsin really are of their work. I think that some of the sentiment data may give mixed results, but, I hope it will show positive and joyful appreciation of the organization.

As I predicted, the sentiment analysis is a bit confusing at times but overwhelmingly positive. Three of the five largest contributors to the sentiment analysis were anticipation, trust, and positive sentiments. These are all good sentiments for the DNR. I think that people ultimately do appreciate the efforts of the DNR. Some of the tweets that might be associated with fear and negative sentiments are likely related to news releases about wildfires in the state.

Fishing is a multimillion dollar industry in Wisconsin ranging tourism, guide services, commercial harvesting, sales related to equipment and more. The state has gained a national spotlight over the last few decades from the strong support from the DNR.

Predictive or Prescriptive Analysis

Hypothesis is that the tweets that have positive sentiments about the organization will have more retweets than those with negative sentimnts like anger, fear, disgust. I will run a linear regression to show me whether or not my hypothesis is wrong by pulling information on my Sentiment Data and tracking retweet count and sentiments.

## 
## Call:
## lm(formula = favorite_count ~ retweet_count + sentiment, data = DNR_Sentiment)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.7944  -0.7084   0.0576   1.1768  13.4528 
## 
## Coefficients:
##                        Estimate Std. Error t value Pr(>|t|)    
## (Intercept)           -0.602241   0.428375  -1.406   0.1603    
## retweet_count          1.885195   0.023812  79.168   <2e-16 ***
## sentimentanticipation  0.891685   0.505866   1.763   0.0785 .  
## sentimentdisgust       0.024982   0.615759   0.041   0.9677    
## sentimentfear         -0.574543   0.485825  -1.183   0.2374    
## sentimentjoy           0.369615   0.576042   0.642   0.5213    
## sentimentnegative      0.109100   0.486704   0.224   0.8227    
## sentimentpositive      0.544652   0.472260   1.153   0.2492    
## sentimentsadness       0.007353   0.548768   0.013   0.9893    
## sentimentsurprise      0.635569   0.595034   1.068   0.2859    
## sentimenttrust         1.149465   0.508595   2.260   0.0242 *  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 2.354 on 611 degrees of freedom
## Multiple R-squared:  0.914,  Adjusted R-squared:  0.9126 
## F-statistic: 649.2 on 10 and 611 DF,  p-value: < 2.2e-16

The linear regression analysis gives us good information on the sentiments regarding the DNR and re-tweet reactions on twitter. It should help us get an understanding of positive and negative sentiments regarding the organization. I think that this can constantly be done, but, the results will change with the different seasons. For example people will be tweeting or the DNR will be tweeting a lot when fishing season opens, or different hunting seasons open and close. The also will tweet a lot during election cycles because they are a government funded organization.

Overall, it helps me answer my hypothesis that there are more retweets for positive sentiments like positive, joy, surprise, trust, etc.

Conclusion

This capstone research project has been incredible. I learned so much about the fisheries in Wisconsin, and specifically Waukesha. I can see now that my home town has done some incredible things with protecting the wildlife population among these lakes.

I learned from my research question, that Rainbow Trout are the most stocked fish in the county. This was seen in the first visualization. I was not expecting this because it’s uncommon for people to boast about their Rainbow Trout experiences in the Southeastern Wisconsin. Everyone usually talks about their Musky, Walleye, or Bass Fishing experiences. But after thinking about it, stocking Rainbow Trout makes sense because they are such a sensitive species and they need extra halp to survive in the conditions we have to offer.

I also learned that fish are more often stocked as yearlings. When I was told about the stocking in the area, I received misinformation that most fish get stocked as fry or fingerlings. We saw in the second bar chart that yearling fish are the fish getting stocked far more than any of their other age classes. I can now use this data to show my other angler friends that they are stocking bigger fish and that we don’t need to necessarily wait a few years after a stocking event fishing a stocked lake.

I was surprised by some of the data as well because I would not expect Northern Pike to be stocked as often as they are. It was always my understanding that they were extremely successful fish in the wild. I will have to do more research to find out why they are not surviving as much in the wild. I have a hypothesis that Northern Pike are harvested more in the winter time, which when they spawn. Therefore, they aren’t spawning at the rates like muskies which are not harvested to eat at all.

Last, to answer part two of my research question. Pewaukee Lake received the most fish in 2020 by a large margin. They recieved 36.5 thousand walleye in that year alone. I am certain that I will keep an eye on this lake in case there are any changes. I am excited to learn more about the stocking efforts when the 2021 information comes out and share it with the broader fishing community. They should learn more about the stocking efforts so they can have positive sentiments regarding the DNR. Even though there are mostly positive sentiments, they may be able to accelerate the appreciation for the group.