My opinion is that if the specimen is collected and stored, all the information should be written in its label. However, it often happens that several fragments/parts of the specimen are also collected and labeled with a lot of information. In those cases, with the goal of saving time, I think that species, number of the specimen on the collection (linking the sample to the specimen label) and identifier name is enough.

Even when all the information about the specimen is inserted on a database, it can happen that it takes some time between the collection and the insertion of data leading to the loss of information from a datasheet or from the memory of the collector. It can also happen that the software has errors or that during the insertion of several specimens, data errors are inserted, specially with coordinates of spatial location. That is why I think that the main specimen or a representative fragment of the specimen should have a complete label.

I think this is great idea.

@Pablo we have cloud storage now and data storage is low cost now that loosing data is not an issue if we manage it with care.

@Pablo and Vijay

The key sentence is "information is to be as close to specimen as possible". I agree! Thus it is not necessary that the very small label, font 5, with a complete info will be attached to the specimen. It can be a printed A4 paper docs, font 12-14, joined with a corresponding set of specimens. A small label just on aq specimen may be reduced to the Latin Name and a Collection Code. I agree with Vijay that the international self-registered database of the Museum specimens will be highly useful, at least for the GBIF ideology.

Not all materials are safe for museum objects. Plus, different objects have different needs in terms of what we conservators refere to as "preventive conservation". If the materials of the label, the paper, the ink, the glue etc etc are not acid-free, this could damage the object. Plus, some label materials attract microorganisms, and information on the labels may be lost, or the microorganisms, insects etc may harm your object. We tend to label different materials in different ways. Could you please be more precise so I can provide some more information?

Regarding databases, computers and digital databases are not ment to last forever. Hard copies, i.e. printed or written should also be kept so that valuable information does not get lost.

Of course you cannot put everything on a label, but at least in your database and books, you could try to include as musch as possible. It's better to record more information and never use it, rather than look for information that could have been recorded but no one ever bothered to.

I remember attending a talk, and the speaker stressed the fact that a museum would keep in its records even the names of the people who recorded the info for each single natural history specimen. This proved to be very useful, as they managed to spot mistakes continuously done by the same person. So when they had to doublecheck the entries for more mistakes, they saved a lot of time that would have to be spent if they were to check the whole database-they just searched for the entries recorded by the person who did the mistakes they had discovered.

I absolutely agree with Pablo: basic information must be written in the specimen label; I mean taxon name, data collection (locality, date, and georeference if possible), collector/s and sample ID. I don't know how it works with animal samples, but for plant collections the herbarium specimens are often studied and also checked and taxonomically reviewed by other researchers or specialists, which can examine them in the museum, but can also view them on loan. So it is important to maintain a stable link between sample and its data. Creating a database allow to provide supplementary information (synonyms, ecological requirements, conservation status, etc.) and photo documentation. If these data are available online, this can be useful for a first check of museum collections and can preserve information from accidental loss and tampering.

This can vary greatly between institutions, but for unconfirmed species of wet collection specimens we use slight variants of the following format (pt 4 font size):

ISOPODA

Piha, Auckland (NZ)

Tai Haruru Walkway

Coastal Native Bush

4 Jan 2006

Col. V W Meyer

close to beach

@Jannifer indecates on the problem, when the arrangement of specimens in collection does not correspond to the numeration order in museums. It is quite usual situation in museums especially in the landscape imitation groups, or during the intensive sorting according to the key morphological characters. In this case the label must be complete as possible. But the taxonomic collection in its not public but scientific (staff) part, has usually the arrangement as in the library: specimens are arranged as numbered, but sometimes they arranged as card index, according the natural classification currently used by taxonomists. From above, I suppose that the quantity (detalisation) of the information on the label may depend on the type of arrangement, because if the arrangement of specimens is numerical, it is easy to keep most of information in the database and printed catalogue, and put only small part (Latin Name and the collection code) on a label.

Volume of information on a label of a specimen may depend on the specimens arrangement in collection.

”the detailed information is to be placed in the the database (synonyms, publications, georeference: longitude and latitude, date of sampling; locality, hosts, ecosystem type, collector, identification taxonomist)”

aha, and a hacker could make to disappear in a second all the information gathered during decades or centuries...

p.s. OK; put those data into data-bases, too, but keep them also on the labels from the museal collections. It is safer in this way.

yes, we also have to use all kinds of back-ups in case of computers and workstations, and in the field we are recording the information on GPS and various other data-loggers... but, beyond of these technical means, one of the surest method is to write the respective details in a notebook, with a pencil.

@Zsolt, yes, safety of the information on the web is a problem.

Measures:

Mirrors of tha databases is normally used in servers

Local Museum network (additional mirror of the museum database only) without outer access

Printed Numeration index with all data for every specimen

Printed Card Index arranged according to the current Classification

@David, thanks for links to papers on fluid collection labelling.

Good advice given in the first paper of 1995 to use the bibliographic databases for the local museum networks (also to produce labels!) if no special collection database software is available

Of course the publication of 1995 is a bit old (nobody uses Paradox databases any more, because of this the international and therefore permanently upgrading collection databases in the Internet are highly advisable)

Some conclusions can be drawn:

"With ever-shrinking budgets and staff levels", we need to

i) save time and expenses of labelling using the most modern technology tools;

ii) use digital technology tools both for back-ups of labels; to develop databases; to save time of labelling; to retrieval of data spread in the collection labels;

iii) labels must be stable for ageing and suitable for scanning and automatic OCR if the database lost or damaged

Thank you friends all for the productive discussion!

I absolutely agree with Pablo, Christina, Zsolt and others: important, basic information on the label and supplementary information in the database only. There is just not enough space on a specimen label to include all the headings and entries in the database. The label may also become confusing if you are not familiar with the design of that database.

@David, thank you for the book reference, it is useful for our ZIN RAS curators

Even though repeating long texts on different labels is tedious, it must be done. A tag must be like a caption in a figure: it must contain all the information needed to correctly place the specimen. And by 'place' I mean placing it in time, space (taxonomical and geographical) and collection. Therefore it must contain the species (or other taxonomic level), date of collection, locality (lat/long AND the datum of this position) and collection number. Other information can be useful, but putting them into a label could make the label too difficult to write.

Other metadata should go into databases, that have details appropriate for each collection. However, I strongly suggest that new databases should adhere to one of the metadata standards out there, like ABCD, DarwinCore, EML or FGDC/CSDGM. I'd suggest going for ABCD/DarwinCore since it's the standard adopted by GBIF. For people that don't know it, go to http://www.tdwg.org/

The more information you can put on a label, the better in my experience. The problem I've seen time and time again is that information entered into a data base is not always accurate or complete. Many collectors I know also keep a log with additional information. Of course I recognize there are limitations.

@Andre, it is an interesting imagination to look at the information on a label as the crossing point of multidimentional space/scale of data. Evidently the items you mentioned are the most important to GBIF standard.

For local museum the unique number and the name of collector are important as well, because these parameters are in the special directories, e.g. in an archive of the filed survey diaries or in the index of the collection section.

@Charles indicates on the potentially important information that cannot be sorted into the database fields, but only placed in a log

Many participants of our discussion think that any database and especially the international ones maybe crashed ocassionally, or by hackers. Thus, specialists consider the label as the most safe back-up.

@Alexander, I agree that the label is THE most important place to keep the information. After all, it accompanies the object that needs to be identified. However, its space is limited, so we have to chose what information can be put there. The idea is that the most fundamental information must be there. Accessory information can/should be kept in other media, such as paper notebooks or electronic databases.

I don't agree that databases are risky because they can be invaded ou lost. The risk is the same for losing a paper notebook or the collection catching fire. Collection curators should have the same kind of care in avoiding both risks, either by fire-safing the collection or adopting good practices for data management.

I suggest taking a look at DataOne best practices:

http://www.dataone.org/all-best-practices

because they managed to summary all the best practices people should have when thinking about preserving data in a long term.