Research Overview:
      This dataset consists of photographs of frog stomach contents from the Los Amigos Biological Field Station in Madre Dios, Peru. These were used for an inter-institutional authentic research experience (ARE) for undergraduate students at the University of Michigan and University of Notre Dame in spring 2021. We created an open-source bowl game competition that challenges students to identify, measure, and compare diet items across vouchered frog specimens. We designed this laboratory activity around six main goals with specific learning objectives:
1.  Authentic research connection: To understand common challenges and opportunities facing scientists who study patterns and processes driving biodiversity on Earth, including problems outside of their exact area of expertise and to which there is no single right answer
2.  Appreciation for the value of biodiversity collections: To recognize the kinds of questions that can be answered with museum specimens, highlighting the excitement of biodiversity discovery as some of the Neotropical invertebrates may be undescribed species photographed for the first time
3.  Understanding the scale of big data projects: To learn data management and scientific workflow skills essential for large-scale datasets.
4.  Identification of unfamiliar organisms: To practice using available keys (dichotomous or otherwise) as identification resources for unknown species.
5.  Collaboration for successful scientific inquiry: To provide an experience in scientific collaboration and networking to solve challenges too big for any one person to solve alone. 
6. Foster a sense of connection among students: To improve student well-being through inter-institutional course collaborations.
       The results of this pilot implementation and more details about this activity are available in the associated publication. These images are made available so that other instructors can implement the same or a similar ARE with their class, either as a standalone or with a partner class at another institution.
      These photos were generated by the lead investigator, Dr. Joanna G. Larson, with significant assistance from, at the time, University of Michigan undergraduates: Leslie Hamar, Nicholas Farrugia, Courtney Whitcher, and Eva Schwarz. This work was funded by a NSF DDIG awarded to Dr. Larson (F046370-084708). The authentic research experience was designed by Dr. Larson, Hayley Crowell, Dr. Lisa Walsh, and Dr. Alison Davis Rabosky. This work has been published as a peer-reviewed research article with the following citation: Larson, J. G., Crowell, H. L., Walsh, L. L., & Davis Rabosky, A. R. (2022). The Batrachian Barf Bowl: An authentic research experience using ecological data from frog diets. Ecology and evolution, 12(7), e9095.


Methods:
      Live frogs were hand caught during nocturnal visual encounter surveys or in funnel or pitfall traps at the Los Amigos Biological Field Station in Madre Dios, Peru. This fieldwork occurred over multiple field expeditions from 2016 to 2018. Stomach contents were collected either through gastric lavage of live animals or dissection of euthanized individuals. These diet samples were stored in 95% ethanol and exported to the University of Michigan Museum of Zoology (UMMZ).
      At the UMMZ, each sample was examined under a dissecting microscope and photographs of prey items were taken using a Leica camera attached to the microscope. Software associated with the camera inserted a scale bar to each photo. Sometimes multiple photos were taken of one prey item from different angles to capture different aspects of morphology. If there were many of the same type of prey item or if the sample contained only fragmentary elements of prey, a group photo was taken for efficiency. Photos were labeled by the field identification code of the frog from which the prey items were collected and organized into folders by frog individual.
      During the undergraduate authentic research experience (ARE) activity that we designed (described at length in the linked research article), students were divided into groups and assigned a group folder of frog diet observations. For each photo, we asked them to identify every prey item to a higher level of taxonomic classification (e.g., class, order), count how many of each prey type was present, and take basic linear measurements of length and width. Each photograph included a scale bar, which students used to calibrate the measurement tool in imageJ, a free software.

File Inventory:
This dataset includes ten folders named Group X, where X is a number between 1 and 10. Within each group folder are 24 subfolders, each of which contains the photographs of stomach contents from an individual frog. Each group folder contains 250-300 photos (1-2GB). Subfolders are not unique to Group folders and may appear more than once in the dataset. This allows for multiple groups to examine the same diet sample and for the class to explore how different observers can have similar or differing results.

Research Permits:
All frog samples were collected under permits from the Peruvian government SERFOR (Servicio Nacional Forestal y de Fauna Silvestre; permit numbers: 029-2016-SERFOR-DGGSPFFS, 405-2016-SERFOR-DGGSPFFS, 116-2017-SERFOR-DGGSPFFS). All animal-related procedures were approved by the University of Michigan Institutional Animal Care and Use Committee (Protocols #PRO00006234 and #PRO00008306).