Computational notebooks as a tool for productivity, transparency, and reproducibility[Bearbeiten]
The current biodiversity crisis, which threatens 1 million species with extinction (IPBES, 2019), poses the greatest challenge for ecologists who need to understand and predict it, and for conservationists
and politicians who need to manage it. Besides their intrinsic eco-evolutionary value, non-human species provide essential services for humans (IPBES, 2019). For example, insects contribute to crop
pollination (Bartomeus & Dicks, 2019) and pest control (Martin, Reineking, Seo, & Steffan-Dewenter, 2015), while forests store carbon that would otherwise contribute to climate change in the atmosphere (Sullivan et al., 2020). Thus, the understanding and control of biodiversity loss is of upmost importance for human life.
During my PhD thesis I have addressed why some species survive longer than others, and more importantly, how managing biological processes and environmental conditions could improve conservation policies to ultimately avoid extinctions. To tackle this, I develop and use computational simulation models that combine theoretical knowledge and empirical data to simulate ecosystems and draw conclusions and forecasts. Simulation models can be rather complex, and thus simplifications and theoretical assumptions are necessary to assure that models are computationally feasible
and results interpretable. Such assumptions must be made clear when reporting the results of models for careful interpretation by the scientific and public audiences.
The debates surrounding the occurrence and consequences of climate change and, even more timely, the COVID-19 pandemic, have shown the need to carefully present complex research ideas, especially to the general public. Having produced this type content before, to great response from non-scientists, I intend to keep doing it for my research for as long as possible. Most importantly,
throughout my studies, I have come to realize that while final results are well communicated through scientific articles, the raw analytical process, so crucial for trusting the results presented, is often
inaccessible. Colleagues before me came to the same conclusion, which prompted an increasing call for reproducibility in Ecology (Culina, Berg, Evans, & Sánchez-Tójar, 2020; Mislan, Heer, & White,
2016). The increasing complexity and speed of development of project-specific analytical methods requires the establishment of quality standards and review processes (Borregaard & Hart, 2016).
The practice of using computational notebooks can fill this gap.
Computational notebooks are documents containing descriptive text with relevant code and results, combined in a narrative order that documents all stages of the research process (Rule et al., 2019). They can be produced with open-source software, such as Emacs or RStudio, and are to be published along the article they relate to. Despite its advantages, it can seem overwhelming to consider producing such document, since it deceivingly appears to constitute additional work to the already demanding scientific publication process. From my experience, this is not the case. Notebooks actually increase productivity by centralizing writing and analysis, which are then reorganized into traditional publication formats (e.g. main text, figure files, and supplementary material). To popularize this mean of open science, I propose to develop a starter kit that facilitates the use and integration of such notebooks in the publication workflow.
Bartomeus, I., & Dicks, L. V. (2019). The need for coordinated transdisciplinary research infrastructures for pollinator conservation and crop pollination resilience. Environmental Research Letters, 14(4), 045017. doi: 10.1088/1748-9326/ab0cb5
Borregaard, M. K., & Hart, E. M. (2016). Towards a more reproducible ecology. Ecography, 39(4), 349-353. doi: 10.1111/ecog.02493
Culina, A., Berg, I. v. d., Evans, S., & Sánchez-Tójar, A. (2020). Low availability of code in ecology: A
call for urgent action. PLOS Biology, 18(7), e3000763. doi: 10.1371/journal.pbio.300076
IPBES. (2019). Chapter 2.2. Status and trends. In E. S. Brondizio, J. Settele,S. Díaz, & H. T. Ngo (Eds.), Global Assessment on Biodiversity and Ecosystem Services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). IPBES Secretariat, Bonn, Germany. Retrieved 2019-06-18, from https://www.ipbes.net/global-assessment-biodiversity-ecosystem-services (Draft chapters)
Martin, E. A., Reineking, B., Seo, B., & Steffan-Dewenter, I. (2015). Pest control of aphids depends on landscape complexity and natural enemy interactions. PeerJ, 3, e1095. doi:10.7717/peerj.1095
Mislan, K. a. S., Heer, J. M., & White, E. P. (2016). Elevating The Status of Code in Ecology. Trends in Ecology & Evolution, 31(1), 4–7. doi: 10.1016/j.tree.2015.11.006
Rule, A., Birmingham, A., Zuniga, C., Altintas, I., Huang, S.-C., Knight, R., . . . Rose, P. W. (2019). Ten simple rules for writing and sharing computational analyses in Jupyter Notebooks. PLOS Computational Biology, 15(7), e1007007. doi:10.1371/journal.pcbi.1007007
Sullivan, M. J. P., Lewis, S. L., Affum-Baffoe, K., Castilho, C., Costa, F., Sanchez, A. C., . . . Phillips, O. L. (2020). Long-term thermal sensitivity of Earth’s tropical forests. Science, 368(6493), 869–874. doi: 10.1126/science.aaw75786
This project addresses an important (and rather common) problem of research transparency and reproducibility in the field of ecology and biodiversity, where mathematical modelling is a go-to research tool. However, the exact computational methodology is often not shared to detail sufficient for reproducibility. Computational notebooks help close this gap by publishing the code, computational outputs and comments. Their use however is not as widespread as it could be, due the perception of the extra work on the part of the researchers preparing a publication. This project uses a two-pronged approach: 1) sharing the specific notebooks along with a scientific publication and 2) development of a "starter kit" that will help other researchers to incorporate this practice into their routine. I find the project proposal relevant to the open science (especially in conjunction with the reproducibility aspect), feasible (the applicant seem to have previous relevant experience that could be used to build on in a reasonable time) and impactful (thanks to the development and promotion of this practice via the "starter kit" in the research community).
A generally convincing proposals with a clear scope and well defined milestones. Yet the main product (documentation of computational notebooks) is a "common product" in that many such documentations already exist and are freely available. As a reviewer I would have liked to read more about what distinguishes the planned documentation of computational notebooks from already existing ones.
- Name: Ludmilla Figueiredo
- Institution: Universität Würzburg
- Kontakt: firstname.lastname@example.org