The Bromeliad Working Group Constitution
The Bromeliad Working Group (BWG): The BWG is an international consortium of
researchers studying the ecology of bromeliads and their associated macro- and micro-faunal
Policy date: This document summarizes policies as agreed to by vote and consensus of the BWG
membership on 23 September 2015. These policies may be discussed and changed at future
meetings as the BWG grows and changes.
Dataset: a collection of bromeliad-level data collected by the same people in the same place in a
discrete period of time. Datasets are normally collected within a single habitat.
Data owner: the person or persons who collected the data. In the case of a student or
postdoctoral fellow, this includes the faculty supervisor. For other datasets collected by more
than two people, the group should nominate two people to be the data owners for the group.
BWG open release date: the date at which the rest of the BWG may use the dataset without
obtaining permission from the original data owner. The data is open only to the BWG, not
necessarily to the general to the public.
Public open release date: the date at which the dataset becomes available for use by the public.
Membership is open to any researcher studying bromeliads and their associated micro- or macro-
faunal food webs, providing the researcher agrees to the policies on data sharing, data use,
authorship and data publication as outlined in this constitution. Additionally, members commit to
respect the differences amongst us, and to act in ways that value collaboration and support over
competition and criticism. The BWG is a non-hierarchical, multi-lingual and international
2 Data Policies
2.1 Using and publishing data
2.1.a The BWG has two guiding principles about data sharing and use:
- The BWG operates on a system of mutual trust. We believe that all BWG members will
benefit from the open sharing of data, and so the expectation is that BWG members will
upload and share their data within the BWG.
- However, we also believe that no individual should ever bear any cost to sharing their
data. Specifically, publication of datasets as part of multi-site BWG studies should never
preempt prior publication by the original data owner. Data owners always reserve the
right to veto the use of their data in any publication.
2.1.b Procedure for asking data-owner to use their data
- Ask once before using the data in a multi-site analysis, ask again before publishing. This
includes asking the author whether or not they want to be involved in other aspects of the
project. The data owner always has a veto over the use or publishing of their data.
- In the case of multiple data owners for a dataset, the two nominated data owners can give
permission on behalf of the others as long as there is evidence of broad consultation and
2.1.c What happens if the data-owner leaves the group, and cannot be contacted, but has not
published the data?
- The data owner is required to enter a "BWG open release date" at the time of entering
their data in the BWG database. If the data owner has not provided a "BWG open release
date", the data will automatically become open to the BWG 5 years after the data has
2.1.d What happens if a student data-owner and their supervisor cannot reach an agreement as to
data release, use and publication?
- This cannot be resolved at the BWG level - conflict between students and supervisors
should instead be addressed by the relevant University.
2.1.e Labs are expected to contribute data before they can use other BWG data.
- Labs that cannot contribute data immediately, you just have to promise that you will once
you have completed the data collection.
2.1.f Principles for contributing data to the database in a timely manner
- It is expected that data will be uploaded to the BWG database within 3 months of
finishing the collection and identification of fauna and analyses of environmental
variables. Entering data in the BWG database will facilitate the data owners own analysis
as they can then easily access BWG information on similar species (e.g. traits, per capita
2.1.g Data input should include: identification of and contact information for the data owner(s);
date of data submission, BWG open release date, Public open release date, and - in the case of
students or postdocs within a lab, the faculty supervisor contact information.
3.1 Guiding principles: We recognize that there is a diversity of publication cultures represented
within the BWG, and that these differences should be respected. In particular, we recognize that
in Latin America the cost of producing data is higher because less money is available than in
Western countries, and that the velocity of individual publications may sometimes be slower. Our policies on authorship
are guided by recognition of these differences, and a desire to provide a system that is mutually beneficial and encourages
the sharing of data.
3.1.a Role of data owners and data managers in authorships
- All data owners who have contributed data to a scientific manuscript will be invited to
join in the authorship of a paper. Such invitations shall not be limited to the provision of
data, but also include an invitation to contribute in the analysis, synthesis and writing
parts of the manuscript. The number of authors originating from a site will be
proportional to the importance of that site's data in the analysis. Data owners may decide
to decline invitations for authorship.
- When database managers have made substantial and essential contributions to the success
of a data analysis project, it may be appropriate to recognize the contributions of these
individuals with authorship. This may include the creators of data compilation and
- In the event that the number of authors exceeds that permissible by journals, the BWG
will collectively decide on the best solution at the time. It may require limiting the
number of data owners per dataset that are given authorship, or simply referring to the
Bromeliad Working Group as an author.
3.1.b Given that papers with many names have highest emphasis on the first and last author, how
do will we acknowledge the contribution of a few authors when there are several data co-
- The primary authors (those with primary responsibility for collating and analyzing the
data and writing the manuscript) should normally be allotted the first few authorship
positions, and potentially the last position. The primary authors can be indicated in the
footnotes, and some journals now allow for multiple lead authors or detailed explanations
of each authors' role. The other authors should be arranged in some sort of systematic
order that is explained in a captional note.
3.1.c Who is the author/owner of data from Bio-Blitz data collections
- In the case of a BioBlitz or similar multi-data owner dataset, it is suggested that two
people are elected to be co-authors. This can be rotated between papers to allow
authorship opportunities for many data collectors.
4 Open Science
4.1.a Guiding principles: Open science should be a goal of all ecologists, because publically
funded science must become the property of the public at some point. However, we recognize
that there are a broad range of equally valid perspectives on when and how data becomes open to
4.1.b Making BWG data public.
- Data owners will be requested to enter a public open release date at the time of uploading
their data. Every 5 years the data manager will collate all datasets that has passed this
public open release date, add appropriate metadata, and place this collated data on a
stable data repository such as - for instance - The Knowledge Network for Biodiversity
or Dryad. In this way, the open-to-the-public database will grow cumulatively at five year
increments. Data owners will be contacted prior to data release for final approval. If
numerous unsuccessful attempts have been made to contact data owners over a six month
period for this final approval, permission is implicitly given to release the data. The open
data will be released under either Creative Commons license CC0 or CCBY following
further discussions within the BWG.
- Lead authors of BWG manuscripts are encouraged to use best practices in reproducible
research and open science, including the publication of related programming code and
publication in open science journals. It is understood that these goals may not be
attainable by all authors for a variety of legitimate reasons.