Journal of Statistical Ecology Journal of Statistical Ecology

Author guidelines

The Journal of Statistical Ecology publishes research at the interface between ecology and statistics and as such all articles must aim to communicate to this dual audience.

Peer-review process

JSE follows a double-blind peer-review process, and articles must be anonymized at submission. Please remove all references allowing to identify authors upon submitting your manuscript.

Article format

At first submission we ask authors to provide a compiled pdf of their work alongside the original source files. The submitted manuscript should be double-spaced with line numbers. Figures should be placed within the text near their planned locations rather than at the end.

The Journal of Statistical Ecology relies on LaTeX as baseline format for its articles, like all other Mersenne centre journals, which promotes cost-efficient and elegant editing of papers with formulas. We therefore encourage authors to use LaTeX from the outset in the production of their manuscript. They do not need to use a specific format at submission, provided that it matches the abovementioned requirements. Please find the LaTeX templates for accepted manuscripts here.

Possible alternative formats before final submission include Quarto/Rmarkdown or Lyx that can be converted directly into LaTeX, and all these tools are free. When writing multi-authored papers, an online software such as Overleaf allows to use LaTeX without installing it on all coauthors' computers. Overleaf has a Rich text mode resembling Word and LibreOffice.

If the authors have written their manuscript using Word or similar (e.g. LibreOffice), and none of the abovementioned options works for them at revision stage, they can provide their source file as a Word document for translation into LaTeX. This possibility is given so as to make JSE as inclusive as possible. However, authors should be mindful that (1) the translation Word-to-LaTeX incurs additional costs to the journal and (2) it currently has to be outsourced out of the Mersenne Centre. JSE therefore reserves its right to modify Word document submission policies at any time to maintain a proper functioning of the journal.

Article length

Although no strict limits are set, articles at Journal of Statistical Ecology should aim to be concise and should be below 20 published pages (in a single-column, single-spaced, 2.5cm margins, default 11pt font size, bibliography and appendices included). The Editors may request shortening of text when they deem a manuscript too long for its content.

Data and computer code

We ask authors to provide all data and computer codes that their results might rely on at manuscript submission, so that it can be examined during the peer-review process. This can take the form of a repository at Github/GitLab or an institutional archive, with anonymization enabled. We require data/code archiving before the manuscript is finally accepted for publication, when anonymity is lifted. Whether full reproducibility of methods can reasonably be achieved is a criterion considered for the eventual acceptance of a manuscript.

Small datasets can be archived as Supplementary Information in .csv format together with the article. Larger datasets and/or code should eventually be archived in repositories such as Zenodo, Software Heritage, OSF, Dryad, or a nationally recognized archive providing a DOI for the data and/or code, together with perennial free access to said data. In cases where data from a previous publication is used (e.g. for illustration purposes), authors may refer to the DOI from this previous publication's dataset.

In a number of instances it may be possible that a dataset cannot be made entirely public (e.g. endangered species location, sensitive state-owned resource exploitation datasets, data pertaining to human subjects, ...). In these situations non-disclosure of data is possible but should be well-justified: the fact that the data may be reused in future publications, for instance, is not reason enough to keep it private, as most repositories allow an embargo period for this specific purpose. In any case, an application of a statistical method on data which cannot be published freely must be accompanied by an application to a lookalike simulated dataset, available freely, to maintain the reproducibility of methods.

Use of artificial intelligence and machine learning tools

Submitted manuscripts should be written by human authors. However, JSE also acknowledges the diversity of the global scientific community, with many potential authors who might not feel fully at ease with their English, and do not necessarily have access to colleagues who can help them improve their text. Therefore, for specific tasks such as translation to English or improving the wording of some sentences, authors may use language models, including large language models (LLMs) and other artifical intelligence (AI)-assisted tools. But these must be used with parcimony and care. Authors should proofread any machine-generated text and are held entirely accountable for its output, including proper citation of sources. The same rules apply to any use of LLMs during coding.

Any use of generative AI must be disclosed during the manuscript submission process in the appropriate box, with the tool used. Failure to disclose use of generative AI may lead to article rejection. Substantial use of generative AI—for LLMs, anything going beyond translation and rewording as sketched in the above paragraph—must also be disclosed in the manuscript Methods section.

Authors may use machine learning/AI methods such as deep neural networks and generative models (e.g. diffusion-based) as a way to analyse their data or develop new machine learning techniques, provided that these are sufficiently described in the Methods to be reproducible, and that they reasonably abide by JSE's rules for Data and computer code (see above).

Any use of AI not already covered in this subsection of the guidelines must be disclosed in the Methods of the article as well, with an indication of the tool used and the tasks performed. Figures cannot be generated directly by an outside generative AI tool that is not fully described in the Methods, as among other issues it impedes reproducibility.

Mathematical style guide

The Journal of Statistical Ecology considers articles that bridge ecology and statistics, with a planned readership in both disciplines, in addition to interdisciplinary scientists. Clear notation is therefore paramount for good communication, and we suggest below a couple of rules to homogeneize mathematical notation.

use capital bold for matrices (e.g. $\mathbf{A}$), as well as simple bold for vectors (e.g. $\mathbf{x}$) in $\mathbb{R}^n$
use $\texttt{\\mathcal\{\}}$ for probability distributions, e.g., $X \sim \mathcal{N}(\mu,\sigma^2)$ for a Normal random variable $X$ with mean $\mu$ and variance $\sigma^2$, unless this generates confusion. If another style e.g. $X \sim \text{Normal}(\mu,\sigma^2)$ is chosen it must be homogeneous throughout.
avoid symbols made of multiple letters (e.g., $AB_t$ to mean adult biomass at time $t$) as this can be confused with a product.
equations are part of sentences and punctuation must be respected. This rule is further explained in this guide on mathematical notation for ecologists.

These writing rules should be viewed as guidelines and can be broken on occasion when the context demands it.