• Why sharing biodiversity Information?
• What to share?
• How to share it in an efficient way?

A first issue will be to determine what the needs of the different biodiversity data end-users are and which are the benefits of sharing biodiversity information.
There is a set of shared data to be agreed upon in order to be enough informative and useful to the different end-users. Needs are numerous, like for example:

• Point to the wanted species
• Point to the wanted specimen
• Do the same species or specimens exist in other collections?
• Where do the species occur?
• Find a group of species or specimens according to their characteristics
• Who is the data provider?
• Who to contact and where to look for additional information?
• ...

Data matching these types of requests should, in my sense, be accessible centrally and exchanged among biodiversity data users in recommended standards. Additional data can be shared and made accessible locally by adding whenever possible appropriate links to the data hold centrally.
The general principle of "how to share biodiversity information" according to GBIF (Global Biodiversity Information Facility) views could be illustrated as follow:

(This is not a schema provided by GBIF but a generalisation of the schema used by the Belgian Biodiversity Information Facility, BeBIF to illustrate its work in Belgium)

Issues and problems, as well from an ethical, scientific and technical point of view, have to be clearly identified, existing and new solutions to be discussed thoroughly and compromises to be found to get things going.
A non-exhaustive list of topics is provided by the organizers of the e-conference:

a. Problems and advantages of sharing information
b. Sharing of data for profit or for free?
c. Are different levels of access acceptable or convenient?
d. How data providers can protect their rights. Are good will and "conditions and provisos for use of content" enough?
e. Should users of Biodiversity information be registered?
f. How to encourage data sharing?
g. It is sensible to expect free access to private collections?

This session seems clearly to concern more ethical issues, which might arise by sharing biodiversity data. A second issue would be to identify the benefits and problems of sharing this kind of information for the different users, being aware that these issues concern as well scientists, decision makers and the general public.

Before entering the heart of discussions, we should remember that we do not live in a perfect world and that all problems cannot be solved, nor all needs be fulfilled. Compromises have to be found. For example it is very important to take care of the technical point of view of IT Specialists. The progress made in computational science, the wide usage of the web, make a lot of end-users believe that identified needs or problems can be solved in a few seconds with a simple mouse click, forgetting that there is a lot of programming behind most functionalities if the expressed fundamental needs for security, efficiency and data quality guaranties want to be fulfilled.

Indeed if some expressed needs are totally justified and interesting from a scientific point of view, they may ask so much computational cost, as well in funds and/or time, that efficiency threatened to get lost. Let's not forget that to share biodiversity information, the tools to share this information have also to be shared. These tools have to be as efficient and low cost as possible which brings as back to the idea that OpenSource technologies should be preferred to commercial ones to guaranty real interoperability, especially with developing countries.

I will now express some ideas coming to my mind when looking at the provided discussions topics:

a. Advantages and problems of sharing information

Advantages:

• Sharing information gives added value to the data, they can be seen and discussed, comments can be submitted by experts and questions can arise. Knowledge gaps can be identified.
• Sharing information enhances collaboration and enables to work in a multidisciplinary context. Scientists can have access to complementary information concerning their data, like for example find out what biosequences already exist for the studied species in reference databases concerning the subject.
• Sharing information in an efficient and adapted way for the different target end-users enables to the different actors to better evaluate the situations and take decisions, elaborate policies, which match better the real needs.
• Sharing information in a open and user-friendly way through the internet may touch a wider public and increase consciousness of the importance of biodiversity related issues.

  Problems:

• There is such a huge mass of data, so how to interlink these data in the most efficient way? Is it possible to have a central access point and track back the wanted information in an easy and quick way?
• Considering this huge amount of data, is it possible to guarantee the accuracy of the data shared and shown online? Are efficient validation tools available and realistic at longer term with the growing data volume? Can we be sure that the end-user will not get lost? Will it not take to much computer time to find the wanted answers?
• How can the data providers rights be efficiently protected, how to guarantee the security and minimize the misuse of the data? Biodiversity data acquire value for publication with time, for example, if sampling is repeated over several years. Data about endangered species or dangerous bacteria strains are collected. Is it safe to share such data, which are important for biodiversity research but where a risk of misuse exists.

b. Sharing of data for profit or for free

GBIF principle is to make biodiversity data freely available to all, adding the issue of profit might spoil the whole idea. Now we of course all agree that collecting such data is as well time and cost effective. There is a prize to pay for human and technical resources. These costs might be paid by public and/or private funds, by the selling of the products issued from the correct usage of the information shared. There should be an agreement on minimal but enough informative data to be accessible freely.

c. Are different levels of access acceptable or convenient?

Different level of access can offer some guarantees for private or sensible data, but requests more complicated IT developments. And how to be sure of the security at longer term? To get things going, a solution might be to focus first on non sensible data and deal with different access levels later, but the basic IT tools used, should be designed in such a way that security can be added afterwards without having to reorganized all the work done.

d. How data providers can protect their rights. Are good will and "conditions and provisos for use of content" enough?

• This is a matter of mutual trust, difficult to assess, which brings us to the previous point of different access level.
• Data providers have also a responsibility here by being aware of the importance of submitting accurate and up to date metadata along with their data. Regular upgrades are necessary. Efficient technologies exist to link each data to its owner(s), but there is nothing we can do to respect IPR properly, if the metadata are not provided and regularly updated.

e. Should users of Biodiversity information be registered?

Yes this can be very interesting to make statistics of usages, give feed-back to end-users and lowers the temptation of an misuse of the data.

f. How to encourage data sharing

• By showing the added value and usefulness with a few good chosen examples.
• By being aware of end-users needs and try to fulfil real needs and gaps.

g. It is sensible to expect free access to private collections?

• Definitely, passionate people often hold private collections and the collections tend unfortunately to get lost or be dispersed.
• Must also show them the added value of sharing the information, getting known, guaranty that their work or possession will not get lost, guaranty that their name will always be associated with the data ...

Only by listing these few ideas and issues it is clear that there is a lot to discuss and more important to do. Tasks must be accomplished in a collaborative way by sharing not only data, but also ideas, experiences and tools. Everybody cannot do every thing at the same time. Priorities have to defined, tasks spread efficiently among the different actors according to their missions and expertise fields. Compromises have to be made and the needs of all the different users understood.