ALL VOLUMES SEARCH TIEE
VOLUME 20 TEACHING ISSUES AND EXPERIMENTS IN ECOLOGY
PRACTICE

Reorganisation following disturbance: multi trait-based methods in R

Clown fish in bleached anemone (Photo by L. Richardson)

AUTHORS

Laura E. Richardson1, Camille Magneville2,3, Laura J. Grange1, Jennifer L. Shepperson1, Martin W. Skov1, Andrew S. Hoey4, Adel Heenan1

1School of Ocean Sciences, Bangor University, UK

2MARBEC, University of Montpellier, France

3Center for Ecological Dynamics in a Novel Biosphere (ECONOVO), Department of Biology, Aarhus University, Denmark

4ARC Centre of Excellence for Coral Reef Studies and College of Science and Engineering, James Cook University, Australia

Corresponding author: Laura E. Richardson (l.richardson@bangor.ac.uk)

THE ECOLOGICAL QUESTION

What effect did a severe coral bleaching event have on the multi-trait structure of coral reef fish assemblages?

FOUR DIMENSIONAL ECOLOGY EDUCATION (4DEE) FRAMEWORK

  • Core Ecological Concepts:
    • Community
    • Biosphere
  • Ecology Practices:
    • Quantitative reasoning and computational thinking
    • Designing and critiquing investigations
    • Communicating and applying ecology
  • Human-Environment Interactions:
    • Human-accelerated environmental change
    • How humans shape and manage resources/ecosystems/the environment
  • Cross-cutting Themes:
    • Systems
    • Spatial & Temporal

WHAT STUDENTS DO

Students use the statistical programming tool R to examine how fish communities are impacted by a disturbance event. Students receive an existing dataset on the abundance of coral reef fish from underwater visual census of belt transects before and after widespread coral bleaching. The dataset contains the counts of fishes from multiple transects surveyed across different habitat types. Students use R, the free software environment for computing statistical analyses and graphics, to subset and pool the data and plot trends in the data at the fish species level. Students select two specific species, research their ecological traits, and relate these to their abundance before and after bleaching. Students then hypothesize how the fish assemblage as a whole in different habitat types might have been impacted by the disturbance event. The students interpret the "functional" trait ordination space after learning about multivariate statistical methods and use univariate statistics to compare trait-based "functional" richness before and after mass coral bleaching. Students relate their findings to the wider literature and summarize their work in a scientific poster that they present to their peers at a student symposium.

STUDENT-ACTIVE APPROACHES

Guided enquiry, problem-based learning, critical thinking, "authentic" assessment (student poster presentation), marking rubric

STUDENT ASSESSMENTS

Student skills are assessed based on their production and presentation of a scientific poster. The poster summarizes the analyses they perform on the dataset in R. The students' ability to interpret and present a scientific evidence-based argument, with reference to the existing literature is assessed in oral and written skills.

The assessment associated with the practical session described herein carries 50% of the module marks and is centred on the creation and presentation of the scientific poster, concerning the results of each individual student's data analysis. The assessment mark is split into two parts: 15% for the oral presentation of the poster and 35% for the quality and content of the poster.

The oral presentation marking criteria include the following categories: 4-minute verbal presentation, demonstrated understanding, structuring, and timekeeping. The structure is expected to include an introduction to the study, aims and hypotheses, methods (data, design, analyses), results, and discussion (see "Marking_criteria_oral_presentation.pdf"). The poster quality assessment is based on the poster content (including the background, aims and hypotheses, methods of data design and analysis, results, discussion, references) and poster visuals (including layout design, graphics, and writing style (see "Poster_quality_rubric.pdf").

CLASS TIME

Total recommended class time for this practical is 16 hours (hr). We split this into six sessions (1 x introductory lecture; 1 x 6-hr practical; 3 x 2-hr practicals; 1 x 3-hr student symposium). The practical sessions can be split as required to enable working through the practical materials. Where instructors do not have enough space in their courses to allocate this time to complete all parts, they might choose to adapt the materials to end the computer-based practicals after Part 2, where they will have computed functional diversity and created the main results figures for their poster (PCoA and boxplot graphs showing fish assemblage trends before and after bleaching). By excluding Part 3, students would miss testing whether observed differences are statistically significant. Similarly, instructors could have students write up their findings (in poster or report format), without extending this to the in-person poster symposium.

COURSE CONTEXT

This practical is designed for third-year undergraduate students, as part of a wider module on "marine ecosystems and processes" in the UK. This equates to junior level on a bachelor degree in the United States. Students will need to have done an "Introduction to R" course ahead of this practical. For example, at Bangor University, all students taking this practical course would have completed module ONS-1001 "Environmental Data and Analysis" in year 1, where they receive training in basic R coding, data wrangling, graphing, common statistical tests and simple linear models, and an introduction to mapping in R. These students then use R to analyse data in several other modules during their first and second year, so would approach this practical with prior experience of some required tasks in R (e.g., creating boxplot graphs, implementing a t-test). If instructors wish to implement this practical with students who do not have any prior background using R, we suggest running two 3-hour workshops where students are introduced to R, covering basic data wrangling, graphing, and statistical tests.

SOURCE

Richardson et al. (2018). Mass coral bleaching causes biotic homogenization of reef fish assemblages. Global change biology 24.7: 3117-3129. https://doi.org/10.1111/gcb.14119. Available free of charge and without subscription here: https://doi.org/10.25903/5b57c26b0beb7 (Chapter 4).

DOWNLOADS

Description of resource files

Student resources

  • Practical Introduction [docx] [pdf]
  • Practical "Part 1": Getting to know the dataset and the ecosystem
  • Practical "Part 2": Building the multi-trait space to measure trait-based diversity
    • Student handout [docx] [pdf]
    • tiny_trait_matrix.csv ## to illustrate how dissimilarity matrix calculation works
    • tr_cat_1.csv ## to illustrate how traits can be categorized into distinct types
    • tr_cat_2.csv ## to categorize traits as distinct types
    • traits.csv ## dataset of species traits from Richardson et al. 2018 supplementary material
  • Practical "Part 3": Analysing the trait-based diversity indices and presenting your data

Instructor resources

  • Introductory overview slides (editable optional PowerPoint file)
  • Practical "preparation" document [docx] [Rmd]
    • This should be updated and shared with students prior to practical to reflect institutional access to RStudio. Currently contains instructions to create an account and use RStudio Cloud.
  • RMarkdown files (part1.Rmd, part2.Rmd and part3.Rmd) and images used to create student handouts for practical "Part 1", "Part 2", and "Part 3".
    • Rmd files include commented-out code that address student questions and has code answers (these do not appear in the student versions).
  • Poster preparation guide (editable optional PowerPoint file)
    • Should be updated with example posters that the instructor considers "good" and "not so good", and to reflect how the instructor wants to run the scientific poster session and receive assignment files.
  • Marking schemes / rubrics for assessment via the scientific symposium

ACKNOWLEDGMENTS

We thank the 2019-2020 Bangor University cohort of OSX-3002 who worked through the first iteration of this practical. Our thanks also to demonstrators Sarah Bond, Helen Ford, Sivajyodee Sannassy Pilly and Tim Jackson-Bue for valuable feedback on the draft material. We thank Nick Graham for part funding data collection, Jacob Eurich and Lizard Island Research Station staff for field support, and Valeriano Parravicini for providing species trait information.

CITATION

Laura E. Richardson, Camille Magneville, Laura J. Grange, Jennifer L. Shepperson, Martin W. Skov, Andrew S. Hoey, and Adel Heenan. June 2024. Reorganisation following disturbance: multi trait-based methods in R. Teaching Issues and Experiments in Ecology, Vol. 20: Practice #3. https://tiee.esa.org/vol/v20/issues/data_sets/richardson/abstract.html

[ TOP ]