Authors
Dr. Jennifer Verdolin
University of Arizona
Objective
Apply time series forecasting and Generalized linear mixed model (GLMM) to evaluate butterfly populations being impacted by climate and land-use changes.
Background
Butterflies are insects that belong to the order Lepidoptera and represent an important group, many of which are pollinators for a wide variety of plants. There are currently about 17,500 described species. Although they occupy many different habitats, all species of butterfly go through four distinct life stages. Because they cannot regulate their own body temperatures, butterflies are sensitive to changes in temperature. Worldwide there has been a precipitous decline in butterfly populations, and scientists are concerned about the possible effects of climate change on these sensitive species. There is already some indication that a warming climate is affecting populations, but the effects may not be uniform for every species. Some species appear to be benefitting from warmer climates and expanding their ranges, while others are declining. Dr. Katy Prudic explains the complicated relationship between temperature and butterfly populations in her podcast interview, Butterflies: The Pandas of the Sky. Despite evidence that a changing climate is leading to population declines, there is still uncertainty over how much influence land-use changes such as increased urbanization or agriculture might also be contributing.
Since butterflies have an incredibly important ecological role in a diversity of ecosystems, population monitoring has been a priority for decades. In addition, there has been a concerted effort in the United States to recruit the public to assist scientists through several key citizen science initiatives. There are those that focus on specific species like the monarch butterfly and others that collect data on all species sightings. In this case study, we will focus on two key citizen science databases: iNaturalist and the North American Butterfly Association (NABA). iNaturalist is a database where thousands of nature enthusiasts contribute species identifications, which are then screened first by a machine learning algorithm and then by at least two human experts. The NABA data are collected by teams of volunteer or community scientists and are centered on tracking butterfly populations over time. The observations reflect the number of butterflies seen by teams during a count. The iNaturalist dataset is a general species-identification database where each observation is submitted individually. The data collected by community scientists are invaluable in helping scientists track butterfly populations. As more researchers utilize these data, however, an important question to consider is how comparable the results are when using different databases. We will explore this by looking specifically at changes in the monarch butterfly over time with data from both iNaturalist and NABA datasets
The Task
Use open-access data for butterflies from iNaturalist to investigate whether there is evidence of a change in the monarch butterfly population size in the western United States between 2001- 2019. These data were filtered to include only data from western states in the US (California, Oregon, Washington, Idaho, Montana, Wyoming, Nevada, Utah, Arizona, New Mexico, and Colorado) and records after 2001 but before 2020 (6,364 records).
We will:
- -Evaluate the change in monarch butterfly abundance over time
- -Evaluate how temperature, precipitation, and urban and agricultural land conversion influence abundance
- -Predict future abundance