Pathways: Starting an Educational Biotic Survey with Moths and Wildflowers

Our goal is to start an Educational Biotic Survey across the continent to

  • teach science and quantitative skills through inquiry-based learning;
  • inventory and monitor a network of field sites to study the impact of weather, invasive species, pollution, pests, diseases and other factors on species and their interactions.

Since 1997 Discover Life ( has had nearly a billion hits and now serves over 300,000 IP addresses monthly. The proposed survey builds on its technology, databases and outreach experience. We propose to

  • expand existing on-line moth and wildflower identification guides and customize them to participating local sites so that species identification is much easier for non-experts;
  • provide training, technical support, and quality control so that a diverse array of participants can generate and use large quantities of accurate data;
  • make these data available via the web within a day of submission so that participants can ask questions of their own data and those of others and learn skills in graphing, mapping, statistical analysis and, for the most advanced, modeling;
  • recruit participants from all walks of life and develop meaningful ways for them to get involved with the survey and learn valuable skills by studying variation among species.

Intellectual Merit
There are two great impediments in engaging the public in studying species: their general lack of identification skills and scientific credibility. You found what? Where? Are you really sure? Observational data that lack voucher specimens or digital verification are plagued by doubt. Did the observer correctly record exactly where and when they saw or heard what they thought they did? Did they identify it correctly?

Discover Life's methods overcome these impediments. Its on-line tools integrate the submission and tagging of digital images, identification, mapping, and oversight by computers and human experts who correct errors and ensure high data quality. Participants take digital photographs and follow rigorous research protocols to document changes in the distribution, phenology, abundance and interactions of species. Ultimately we will support studies of pollinators, lichens and many other groups.

Broader Impacts
We are implementing a completely new and radically more efficient way to collect field data. We plan for our methods to become widely accepted and develop into a dense network of study sites to better understand and manage our environment. They will help teach science and quantitative skills in many educational settings, both formal and informal.

Discover Life helps us to recruit participants and disseminate results. It shares data with the Encyclopedia of Life ( and other sites by providing over 400,000 maps as a web service. These data will enable local site managers to respond rapidly to invasive species and other threats. Our customized identification guides will overcome the taxonomic impediment that hinders much ecological research and management.

We work with many groups and individuals. In Georgia these include K-12 teachers, Georgia Perimeter College, the University of Georgia's Young Scholars, Young Dawgs and 4-H programs, and nature centers and parks. Nationally our partners include the Organization of Biological Field Stations. We actively recruit participants from under-served communities.

Title Pathways: Starting an Educational Biotic Survey with Moths and Wildflowers

Investigators John Pickering, Leah Larkin

1. Project Rationale
In this Pathways project we propose to start an Educational Biotic Survey at field sites across the continent. The survey has twin goals that combine education and original scientific research. These are to

  • teach science and quantitative skills through inquiry-based learning;
  • inventory and monitor field sites to study the impact of weather, invasive species, pollution, pests, diseases and other factors on species and their interactions.

Educationally, the survey will involve students and the general public in every aspect of research, from developing hypotheses to data collection, analysis and presentation of results. In addition to the scientific knowledge and technical skills that participants will acquire, we hope that their exposure to the vast diversity of life around them and the thrill of discovery will inspire many to make the study of science and nature a life-long passion.

Scientifically, understanding the potential impacts of climate change, invasive species, and other large-scale factors on biological systems is a formidable task. It is not feasible to conduct randomized, replicated experiments at regional and continental scales. Instead we can use natural field experiments to study such phenomena (Hargrove and Pickering 1992). The survey's geographic scale and replication of methods across study sites will allow participants to take advantage of natural experiments, such as how weather patterns affect organisms.

This proposal builds on work that Discover Life has accomplished. We propose to start the Educational Biotic Survey by building and testing an infrastructure to study moths and wildflowers across the continent. As detailed below, this entails recruiting a diverse array of participants to inventory and monitor sites, building local identification guides to these sites, and providing training and support to participants. The proposed evaluation will be critical in designing a 'Full-scale Development' proposal, in which we intend to expand to a denser network of sites and a broader array of study organisms.

2. Project Design

  • 2.1 Infrastructure
    The Educational Biotic Survey's infrastructure is based on technology and research protocols developed by Discover Life and its partners. This website was started in 1997 to provide identification guides and other technical support to the All Taxa Biotic Inventory (ATBI) of Great Smoky Mountains National Park, an ongoing effort by scientists and volunteers to inventory the park's estimated 100,000 species. The website's mission has since expanded -- to assemble and share knowledge in order to improve education, health, agriculture, economic development, and conservation throughout the world. With contributions from the American Museum of Natural History, Global Biological Information Facility, Missouri Botanical Garden, Smithsonian Institution, and many university and other contributors, the site now serves information on over 1.1 million species, 100 million specimen records, 400,000 distribution maps, 370,000 images from 500 contributors, and 560 identification guides and checklists maintained by experts. The sites strengths include diverse groups: ants, bees, butterflies, caterpillars, ferns, fish, frogs, ladybugs, lichens, moths, salamanders, slime molds, snakes, ticks, trees and wildflowers.

    We propose to start the Educational Biotic Survey with moths and wildflowers. These groups are biologically diverse, ecologically important, and easily and safely studied by novices. In the past several years we have experimented with several research protocols, including Mothing and Nature Walk, the two that we propose to develop further here.

  • 2.2 Methods
    Discover Life's research protocols and technical advances enable anyone to collect high-quality digital observations, identify species, and analyze information about organisms across geographical and temporal scales. Using digital photography, participants will document the phenology, distribution, abundance and interactions of plants, insects and other organisms. Participants take photographs of a cell phone and GPS unit or landmark to record exactly where and when they recorded data. They then take photographs of species and their interactions, such as a pollinator or herbivore on a plant. Finally, they upload all photographs to Discover Life and tag them as to where and when they were taken.

  • 2.3 Identification
    With nearly 40,000 plant and over 11,800 moth species in North America north of Mexico, the identification task is initially daunting. However, it is much easier if one uses an identification guide that is customized locally. When testing our guide to 500 moths in Clarke County, Georgia, students identifed over 80% of moths accurately in a matter of minutes.

    For all participating sites, we propose to build customized county-level guides to moths and wildflowers. As the final step in identification, our experts will check and correct all participants' determinations. While we cannot identify some species from photographs, our experience in Georgia suggests that we can identify over 90% of moths to species.

  • 2.4 Results
    Mothing is our most advanced protocol. It started in 2010 and has generated over 40,000 photos. High school and college interns, Discover Life's staff, and other participants have identified the moths in 14,000 photographs to 528 species, for which there is now a local guide. Automated computer programs tabulate the most recent results nightly and update them on the web in a set of tables (see Thus, we make all photographs, identifications, and associated data available within a day of submission. We will do the same for wildflowers and make additional tables available, such as daily weather data from the nearest NOAA weather station to each site.

  • 2.5 Analysis
    The data collected by Mothing already supports inquiry-based learning and teaching quantitative methods to high school through graduate level students. Participants can compare their own data with those of others, learning graphing, statistical analysis and, for the most advanced students, modeling. We provide faculty and teachers suggested questions that they and their students can answer. These assignments range in difficulty. They start with simple graphing for high school biology and environmental science classes. The more complex questions should probably only be attempted by college seniors and graduate students with a bent for computer programming and fitting data to biological and statistical models.

    Potential Mothing assignments:

    1. Graph the number of moths photographed in each month from Table 1. Which month of the year has the greatest diversity? Is this the same across years?
    2. Based on Table 2, which moth species have the longest flight period? Are these the most common species based on the number of photographs taken?
    3. Make a pie chart comparing the species at your site with those collected during the same months at the site in Clarke County, Georgia (Table 2). Have the chart represent the proportions of species at both sites, at just your site, and just in Clarke County. Discuss how the following might explain your results: (1) differences in sampling effort and number of moths photographed between the sites, (2) local differences in the habitat and plant communities surrounding the sites, (3) the geographical location of the sites with regard to possible regional, latitudinal and elevational differences, (4) local weather patterns at the two sites, and (5) other factors that might have affected the results.
    4. In terms of the number of moth species photographed each night, which phase of the moon has the greatest diversity? Does this differ across months? Table 4 has the lunar cycle; tabulate the number of species recorded each night from the individual records.
    5. Graph the seasonality (phenology) of a species, comparing it across years (or across sites when we get time-series data from additional sites)? Is the species flying during the same period across years (sites)? Explain what might cause the species seasonality to differ across years and sites. How might Atlanta's "heat-island" affect moth seasonality relative to moths at generally cooler, rural sites such as in neighboring Clarke County?
    6. Choose the four most photographed species (say species A, B, C, D) that flew in April, 2010. Make a time-series graph to compare when they were photographed in April and May in 2010 versus 2011. Did species A, B, C, and D fly earlier, later, or on the same dates in 2011 as compared with 2010? Did the relative order of their (first, median, mode (peak or peaks), mean and last) recorded dates change across years or were they the same? For example, if species A's peak flew before species B's in 2010 did it do so in 2011, or did B fly before A? If the order were ABCD in 2010 and DCBA in 2011, then there would be a complete reversal in the species phenology. Why might this happen?
    7. And for the serious number crunchers, use additional site-year comparisons to model how temperature, rainfall, photo period, lunar cycle, and latitude predict the date when a species will start flying each year and the date it will stop flying. How does your model incorporate the affect of temperature on generation time? If there were an average increase of 5 degrees Centigrade throughout the year, would it predict an addition generation, assuming that the moth's host plants were available?
    8. Why are geometrids apparently more common than noctuids? Answer: because they are easier to identify! Which raises all the issues of data quality, assumptions, and scientific ethics, which everyone should understand. This is real science. Learn to think, question, and give thoughtful, creative answers.
    9. What would you ask of these data? Your questions are the beginning of independent research, discovery, and a wonderful life solving mysteries.

    In a similar way, we propose to post wildflower data so that students can compare, contrast, and explain insect and plant phenologies across sites and years. Consider how insect versus plant development times and activity levels may respond differently to photoperiod and temperature regimes. Under what circumstances is it most likely that climate changes could affect pollination efficiency by disrupting the seasonal synchrony between flowers and their pollinators?

  • 2.6 Recruitment
    We will recruit participants from schools, community colleges, colleges and universities, clubs, nature centers and museums, public lands, and individual naturalists. We have built partnerships with minority recruitment programs such as the Young Scholars in Georgia and the Ecological Society of America's SEEDS program. We will actively continue this commitment to diversity. For citizen science projects such as the Lost Ladybug Project with Cornell University and Bee Hunt with the American Museum of Natural History, we have recruited and trained young people through Chattahoochee Nature Center, Arbor Day Foundation, Environmental Education Association of Georgia, and other environmental education organizations. We have reached thousands of middle school and high school students through NSTA webinar and webcasts, National 4-H Congress, Teacher Quality workshops, and high school summer research experiences. We recruited over 40 biological field stations across the continent as sites (see In addition, many enthusiastic naturalists who post images on Flickr and other websites join Discover Life to use our mapper, species identification guides, life list and other tools. Through social networking, we will enlist these naturalists as mentors.

    We will evaluate diverse types of field sites, whose participants have different motivations to join the survey. We propose to work with at least two sites of each of the following:

    1. High school teachers mentoring students with science fair projects
    2. High school teachers establishing an after school Discover Life club
    3. Middle school 4-H clubs
    4. High school 4-H clubs
    5. 4-H educational centers
    6. Boy Scouts or Girl Scouts
    7. Community college faculty
    8. College or university
    9. High school and college interns
    10. Nature centers, herbaria and natural history museums
    11. Public parks and other lands using the survey for public education and land management
    12. Field stations for public outreach and land management
    13. Natural history societies, associations and clubs
    14. Individual naturalists and nature photographers

3. Project Management

  • 3.1 Organizational structure and governance
    The Polistes Foundation ( is a 501-c-3 non-profit organization based in Massachusetts. It is the legal and fiduciary umbrella of Discover Life. Polistes will administer this grant virtually, overseeing finances, providing administrative support, liability insurance, and fiduciuary reports and audits as required by the NSF. Polistes charges 10% indirect costs.
  • 3.2 Governance -- Executive Committee
    Within the legal framework of the 501-c-3, PIs Pickering and Larkin will serve on the project's Executive Committee along with three additional individuals selected from among the 75 researchers at the International Center for Public Health and Environmental Research, PHER, founded in 2007 ( To minimize travel, they will coordinate their activities via regular Skype conference calls.
  • 3.3 Personnel
    • Dr. John Pickering, PI and founder of Discover Life, will oversee the project, manage the computational infrastructure, and contribute to insect guides. He has a Ph.D. in Biology and over 30 years experience programming computers.
    • Dr. Leah Larkin, CO-PI, has expertise in both botany and entomology. She will recruit colleges and NGOs to participate in California and other western states, will supervise two undergraduates in the creation of wildflower and moth guides for the western United States, and will provide taxonomic oversight for these groups nationwide.
    • Nancy Lowe, Discover Life's Outreach Coordinator, will recruit participants. She will make training videos, coordinate activities, provide technical support, and conduct the evaluation.
    • Steve Bowling has over 50 years experience identifying plants and is a regional expert. He will work with the CO-PI to enhance the wildflower guides, customize them to field sites, and ensure data quality by correcting identification errors submitted by participants.
    • Mary Doll, a recent graduate from the University of Georgia, will do the same for moths. This year she has identified over 300 moth species that she has added to our guides.
    • Dr. Justin Long has worked for Discover Life part-time for 8 years. He provides computer support and is responsible for all systems work, security patches, and backups.
  • 3.4 Scheduled work plan
    • Recruitment of up to 50 sites (starting August, 2012)
    • Training videos and webpages (August-December, 2012)
    • Guide building and testing (August, 2012 - July, 2014)
    • Technical and identification support (August, 2012 - July, 2014)
    • Evaluation (Develop: summer, 2013; Test and analyze: summer, 2013 to July, 2014)

4. Project Evaluation

  • 4.1 Overview
    We propose to conduct an internal formative evaluation, using a quantitative case study. Although this evaluation method will not compare project participants (treatment) to non-participants (control) nor measure changes over time, it will enable us to compare results from different demographic groups to determine which groups benefit the most from our project and how we can better serve those that experienced least success. For example, we may measure higher levels of both knowledge and engagement among community college students than high school students, or we may find that visitors to parks and nature centers are the most enthusiastic participants. These findings will inform future project design, including location and type of future study sites.

  • 4.2. Logic Model
    Please see figure.

  • 4.3 Measurement methods
    We will use a variety of different instruments to measure outputs. We will build web tools to track data such as the number, demographic, and location of participants. One such tool that we have is the "life list" feature, which tabulates the number of species identified by each participant. It also generates competition and excitement through social networking. When participants add new species records to their local checklist, these new occurrences will be validated by taxonomic experts, then added to a list of new records. Each site leader (teacher, nature center staff, park ranger, etc.) will report the number of science fair and other independent projects that participants complete. Discover Life web logs enable us to determine the number of hits to individual web pages such as data tables. To measure the success of our online identification guides, we will assign random groups of moths or wildflowers to participants and determine the percentage of correct identifications. We will analyze species identification success among various age groups and experience levels.

    Using a quantitative case study, we will design and implement two types of post-test for our participants. One test will be for site leaders and individual naturalists. This test will assess the leaders' own experience in participating as well as the overall success of their site. The second type of test will evaluate the experience of participants at group sites such as students, visitors to nature centers or club members. We will compare the results of individuals across different types of sites. What age group reported the greatest shift in engagement and interest? What experience level reported the greatest gains in knowledge and understanding? What demographic had the most trouble with new skills specific to the survey?