Guidelines for collaborations, publications and datasharing

The purpose of these guidelines is to clarify the rules for data sharing and for publications arising from the OPHEE experiment. They are based on similar policies from other projects (BIOTREE, FunDivEUROPE, NutNet…). Actually, they are more than inspired from and borrow heavily from those projects and we must acnkowledge the intial work and phrasing of Alain Paquette, Christian Messier, Michael Scherer-Lorenzen and Elisabeth Borer and collaborators (Borer et al. 2013, Methods in Ecology and Evolution, doi: 10.1111/2041-210X.12125).

The ORPHEE experiment, as member of the Tree Diversity Network (TreeDivNet) was initially planned by Hervé Jactel and is now represented by himself and Bastien Castagneyrol and Céline Meredieu as Scientific Coordinators (SC). The SC are responsible for the coordination of collaborations within ORPHEE and those involving ORPHEE and other TreeDivNet partner experiments.

As SC, we are very happy to welcome scientists wanting to address specific questions on the ORPHEE experiment. We are also happy to contribute to larger studies and meta-analyses by sharing existing data. Because such data relies on many different contributors and contributions, we want to make sure that each of them is duely and fairly acknowledged. We have attempted to lay out ground rules to establish what we believes deserves authorship, or not. We want these rules to be as inclusive as possible while not diluting the value of authorship on a manuscript.

Start an experiment or carry out research at ORPHEE

Proposal

To favour collaborations and avoid conflicts, all projects within ORPHEE should be made public to SC who will take care to inform local managers and collaborators, plus TreeDivNet partners if applicable.

Project proposals should be one page length and should indicate scientific context, research question or hypotheses, methodology and timelines, collaborators, and, if applicable, funding sources. Project proposals will be listed in this website.

This step is very important and should deserve attention. It is meant to inform others and avoid learning too late of an opportunity to collaborate on a project. This is the ideal time for members to inform the leaders of a new project of their interest to collaborate (see below for authorship issues).

As soon as data has been acquirred and the experiment is considered completed, all scientific equipments, small and large materials must be removed from the field.

Specific requirements

Items listed below are not compulsory but need to be discussed with SC:

• All projects within ORPHEE must use protocols common to other TreeDivNet experiments when available, or make one up for the future. This is for consistency to insure that multisite analyses will be possible. New collaborators agree to make there protocoles publicly available to future collaborators.
• Destructive samplings in particular should be well thought of and imperatively discussed with all those that may be affected.
• When planning their sampling campaigns, teams should always keep in mind the best interests of the ORPHEE and TreeDivNet community. For example sampling efforts should be planned in order to be useful to as many projects as possible in order to: (i) Reduce unnecessary destructive sampling and disturbance by combining measurements; (ii) Maximise analysis capacity by linking different measurements made at the same locations. It belongs to SC to connect applicants with other collaborators who might have immediate or future interest in particular samples.

Data sharing

Data production, archiving and use

As an a priori rule, data collected within ORPHEE will, within reasonable limits detailed below, be available to scientists and collaborative use of data will be favoured. Collaborative synthesis does complement rather than forestall disciplinary publications. However exceptions are possible, for example with sensible data.

Data collected within ORPHEE should be deposited into a central database (managed by the SC), after they have been quality checked by the PI of the subproject delivering the data. Data need to be deposited as quickly as possible, normally within one year, and at the latest two years after the field sampling or laboratory analysis has been completed.

This data will not be made publicly available. However, every new participant will need to provide meta-data that will be made publicly available. At least, meta-data must indicate rationale, contact person, collection date, sampling level (block, plot, tree, leaf) and sampling effort.

Whenever possible, data collected should be shared among all other members to favour overarching analyses. However the scientists originally generating the data must be informed about any planned use of the data, and the use of data is not allowed without their approval. Scientists making use of data supplied by others for publication must acknowledge the use of the data appropriately (see below). In principle, data from other projects shall only be used as covariates or additional information, but may not be analyzed in the way originally planned.

The form of credit to people who collected the data should be discussed before a paper is written. Joint co-authorship should be the standard if data from other researchers contribute significantly to the main findings. For any release of parts of the dataset to people that are not members of OPRHEE or TreeDivNet, permission has to be sought from the SC and the PI of the project that generated the data.

Existing data

Some data was used in scientific publications, some is still unpublished (yet). If applicable, you will be requested to contact corresponding author(s) of related papers in addition to people listed below:

– Insect herbivory: contact Bastien Castagneyrol (bastien.castagneyrol@inra.fr)

– Dendrometry (tree height, dbh): contact Celine Meredieu (celine.meredieu@inra.fr)

– Understorey vegetation: contact Bastien Castagneyrol (bastien.castagneyrol@inra.fr)

– Arthropod diversity: contact Hervé Jactel (herve.jactel@inra.fr)

– Soil properties, tree nutrition and biogeochemical processes: contact Laurent Augusto (laurent.augusto@inra.fr)

Authorship, acknowledgements and related issues

Whatever author list, It is compulsory that ORPHEE is listed among keywords. Likewise, each paper using data from ORPHEE should acknowledge UE Forêt-Pierroton:

"Authors acknowledge UEFP 0570, INRA, 69 route d’Arcachon, 33612, CESTASEUFP-0570 for the management of the ORPHEE experiment".

And each paper in which irrigation was considered as an experimental treatment should acknowledge the EquipEx Xyloforest: "Grants from Investissements d’avenir, Convention attributive d’aide EquipEx Xyloforest ANR-10-EQPX-16-01"

Given modern pressure on publication, we acknowledge this is a sensible issue that must be dealt with early in the process leading to a paper. It is imperative for ORPHEE that results obtained will be published quickly. The careers of the researchers, students, and postdocs involved as well as the chances for continued funding depend on rapid publication.

The planning of an experiment, its design and hypotheses, the planning and executing of data collection, the analysis of the data, and the writing of a manuscript all represent necessary steps in producing a paper that can merit authorship.

In short, authorship is based on whether the person has contributed to at least some of the following stages: (i) planning and establishment of the experiment, (ii) collection of data, (iii) analysis of data, and (iv) interpretation and writing.

Authorship is something that is deserved and earned based on the importance of the intellectual contribution to any one paper. The present agreements in no way change that. The planning of the project constitutes an essential intellectual basis for the success of all projects within ORPHEE. In addition, the planning, set-up and maintenance of the experiments do not deliver any data to the scientists mainly involved in these activities. Although, as soon as one experiment stands alone, lives its own life, and has been described in many previous papers, whether having been involved in initial planning deserves authorship is debatable. This question is not to be solved here and, as a general rule, we ask that SC are offered authorship. They may estimate that their contribution is not strong enough and may ask to be simply acknowledged.

The first and most important step in achieving these objectives and avoiding unecessary conflicts is the early sharing of projects. As soon as a team has discussed research questions, hypotheses and methods (i.e. a research proposal, for example for a grad student), they should communicate that to the ORPHEE community and list the project up on this document. The proposal should include the list of people involved. This step serves many purposes:

• Inform other members of the project’s protocol, so as to make sampling more efficient collectively (avoid measuring the same things, or destructive samplings that could have been combined), and the better use of data
• Favour collaborations and multi-site projects. Collaborators from other sites may offer to participate in a new project by using the same protocol at their experiment, or facilities for students to come on their site, as we have often seen already
• It constitutes an invitation to contribute. This is the best time for others to let the proposers know of their interest to participate in the new study and make a significant contribution. However this does not mean the researchers involved have to accept.
• Finally, this also serves to verify early that people that deserve recognition (e.g. through authorship), but may have been forgotten, are given a chance to contact the researchers involved early in the process, so as to avoid frustrations on all parties.

An invitation to collaborate does not constitute automatic attribution of authorship. A significant contribution is still required for that. Only the invitation allows the person invited to decide for themselves whether they would like to contribute or not.

Choosing to contribute comes with responsibilities on the part on the contributor as well:

• To avoid slowing down the publication of results, we ask that contributors provide their comments and interests in a project as soon as can be arranged
• Failure to provide a significant contribution within reasonable time may result in the person being removed from the pool of co-author.

As a last step, all manuscripts meant for publication should be distributed, at least 30 days before they are submitted first, to SC and ORPHEE collaborators. This step thus only serves to inform and allow for final improvements if offered (not assuming authorship in return, i.e. friendly reviews), as well as making sure no one was forgotten. We encourage researchers to share manuscripts at earlier stages of writing (when e.g. results are complete) to seek opinions and comments, and possible collaborations.