Turtle Pond, Pepperwood Preserve, Sonoma, CA
Labeled Milkweed Plants
This past year I have been exploring Monarch Butterfly ecology. Though my project has shifted multiple times I have maintained a steady focus on Monarch Butterfly behavior and the possible influences of their dramatic population decline. Through this exploration I have connected with my mentor, Dr. Louie Yang at UC Davis. I have also had the opportunity to collaborate with one of Dr. Yang's grad students, Dylan MacArthur-Waltz. Dylan is exploring different facets of milkweed monarch interactions and is collaborating with Pepperwood Perserve in Sonoma California to monitor natural milkweed populations. Over the past two months I have visited the Perserve every other weekend in order to record data on the natural population growth. In the coming weeks I aim to determine my independent project and begin working towards a finalized project proposal.
This week I started learning R!
Since I last updated this website a lot has changed. I have narrowed down my project focus and shifted my research collection from in-person, to citizen science. For my project, I will be collecting iNaturalist data on milkweed observations in metro areas around the US. Then I will be using R to analyze this data. Exactly how, I'm yet to determine.
But, that means I'm learning R.
Pepperwood Milkweed Plants
On September 11 I visited the Pepperwood Perserve again. Last spring I was visiting Pepperwood pretty regularly with visits most Sundays but during the summer these visits became less frequent. This Fall we did one final visit in order to see how the milkweed population has changed over the summer. During this visit we discovered that the location we were observing for milkweed plants was actually not the site where the most milkweeds were growing.
Ultimately this visit was mostly useful for future projects and not applicable to my own project.
This week we started to analyze the iNaturalist milkweed data and begin to collect results of this project. The goal of my project is to determine the abundance of tropical milkweed in major metro areas around the united states over the recent years.
Over the past few weeks I have gained a greater understanding of R through the R learning course provided to me by my mentor and through basic data setup with Dylan.
iNat California Milkweed Dataset
Number of Total Observations over time
Proportion of Observations which are Tropical Milkweed in the Bay Area
I have made a lot of progress on my R analysis of the iNaturalist data. Here are a couple things I have done:
Created city by city tropical milkweed proportion over time graphs
2. Visualized the trends of all milkweed species in a metro area overtime
3. Began doing pairwise proportion tests to analyze the if metro areas within a state are significantly different from each other
Over the past month I have made a lot of progress. Through my weekly meetings with my Mentor, Dylan, I have been able to do all the needed statistical tests. At this point, or code for the project is almost complete and when we return in the New Year I will begin the writing/presenting process. The statistical tests took me a little while to understand but here is what we did. In R, we used the pairwise proportion test function in order to compare a number of different groups. The first comparison occurred between states. I used the entirety of the state data and compared the proportion of milkweed that is tropical milkweed to each other. Essentially the test looked at the probability (p-value) that the long-run proportion of tropical to non-tropical milkweed is the same between the two locations. So for the state-to-state comparisons, it finds the probability that the long-run proportion of milkweed data that is tropical milkweed in California would be the same as the long-run proportion of milkweed data that is tropical in Texas or California to Florida or Florida to Texas.
In addition to the state to state comparison, I also used the pairwise proportion test to compare metro areas within the states to other metro areas within the same state. And, I used the same test to compare the years in each metro area to other years in the same metro area. I am then going to use the resulting p-values to determine the trends over time and analyze different states and metro areas.
See Screenshots of the Results from the Code Below
At the moment I am trying to enter the Golden Gate STEM fair. Though this process has been difficult, I am aiming to have a significant portion of my results analyzed so that I can pull together a presentation for March. So far in this project I have began to think about the framing of my results and edited down my intentions to the following abstract.
Project Abstract:
Using Citizen Science Data to Determine Long-Term Abundance Patterns of Asclepias curassavica in US Metro Areas
Monarch butterflies are highly reliant on their host plant, milkweed (genus Asclepias). One milkweed species, Asclepias curassavica (native to central and South America) has been introduced to the United States by the horticulture business because it is simple to grow and ornamental. Its effects are much debated. A. curassavica is a nearly year-round stable food source for monarchs and potentially one of the causes of growing non-migratory populations. Researchers are unsure that non-migratory populations can sustain the species long term, although A. curassavica appears to offer some potential to maintain these populations. Conclusive evidence suggests that resident populations have higher levels of Ophryocystis elektroscirrha (OE), a protozoan parasite. In non-migratory populations, a lack of winter senescence means OE doesn’t get cleared. My research addresses the question: How has the abundance of Asclepias curassavica plants in US metro areas (SF Bay Area, Los Angeles, San Diego, Austin, San Antonio, Houston, Dallas and Fort Worth, Orlando, Tampa, Miami) changed in recent years? I hypothesize that the abundance of tropical milkweed plants will have increased because of the popularity of monarch conservation and other social campaigns that possibly inappropriately encourage the planting of this species. To answer this question, I have collected iNaturalist data on milkweed plants (genus Asclepias) and used R to conduct pairwise statistical tests to compare the proportion of milkweed observations that are tropical across years, states, and metro areas.
I am very excited to continue working on my project. For the next month I am going to focus all my attention of finalizing my code, collecting any last data, and creating a solid analysis which I prepared to present.
This month I completed my STEM fair entry. For this, I worked through a full abstract, introduction, methods, results, and conclusions. Below is my project poster and project booklet.
I won my division (12th-grade biological sciences) in the Golden Gate STEM fair. I also won the National Geographic That's Geography award and the California Native Plant Society Biodiversity Award. For this fair, I interviewed with a panel of judges virtually and then went in person to the Sausalito Bay Model to receive awards.
Image credit: Getty Images