Towards a digital image collection of the Herbarium of the University of Granada

M.Teresa Vizoso

The application of new technologies in the field of natural-history collections is enabling herbaria not only to manage their collections more effectively but also to provide access to a large volume of biodiversity information through computerization and the creation of a specimen database. After this step, databases can enrich their contents by digitizing images associated with high-priority specimens such as nomenclatural types. This represents a major advance in the preservation of sensitive materials. For example, the risk of damage involved in consulting and sending sensitive material is minimized. It is also vital in terms of enhancing information and reaching out to wider audiences at multiple levels, since, at the database level, database records can have links to the associated image (Häuser et al. 2005). At a more general level, institutional, national (e.g. GBIF.ES) or global (e.g., Encyclopedia of Life) species catalogues can combine this information with their own and provide it to multiple audiences. In this project, the GDA Herbarium has sought to respond mainly to the increasing demand for the use of the images in order to preserve sensitive material. The herbarium started to generate digital images of three of its most sensitive collections: nomenclatural types, the historical collection of Amo y Mora (1852) (the oldest collection of this herbarium) and other digital images of sheets from 200 Andalusian singular taxa that illustrated the "Singular Flora from Andalusia" species portal, also developed by the herbarium. The techniques that allow the incorporation, operation, and management of new digital-image collections, as well as those which facilitate its accessibility via Internet have also been implemented.

The main objective was to ensure the preservation of sensitive material of the University of Granada Herbarium by implementing high-quality curatorial standards as well as to generate, manage, and disseminate the associated data in digital formats. This was achieved by:

• Creating and making available on Internet an image collection of singular taxa from Andalusia that meets the needs of researchers, managers, and the general public.
• Obtaining a detailed report on the number, location, and treatment of nomenclatural types in the University of Granada Herbarium in order to perform a database-cleaning process and to complete the curatorial management of this sensitive material.
• Compiling a collection of electronic publications (protologues) that support the study of nomenclatural types and improves accessibility and service that the University of Granada Herbarium offers to researchers.
• Ensuring the preservation of the historic collection of Amo y Mora (the oldest collection of this herbarium, started in 1852 by M. Amo y Mora) by fully digitizing it and publishing it online.

The Herbarium of the University of Granada is part of the research group called “The Herbarium of the University of Granada as a Source for Taxonomic, Environmental and Biodiversity Studies” (group code: 288RNM) which is one the Research and Technological Development groups from Regional Government of Andalusia (Junta de Andalucía), Spain. The funds were provided by the Technical Support Program, under the Scientific-Technological Infrastructure modality of the Ministry of Science and Innovation of Spanish Goverment and the University of Granada. The project was undertaken from February 2010 to July 2013.

This study includes fungus and lichen collections. The fungus collection of the herbarium comes from two herbaria, one of which originated at the Pharmacy Faculty (GDA) and the other at the Science Faculty (GDAC). In the year 2000, the two herbaria were joined. The Herbarium of the University of Granada (GDA) maintains both sections and, although both collections retain their individuality, new materials are added only to the GDA collection. Currently, the overall fungus collection comprises 7244 specimens derived mostly from research conducted by Dr Antonio Ortega. Some 97% of the specimens come from Spain, followed by just under 2% that correspond mostly to exsiccate Uredineen Sydow (1874-1889) of a collection of Uredinales, Ustilaginales, and Erysiphales, most from Germany and a small portion from other countries such as Hungary, Austria, and Sweden. There is also a small number from Portugal and other Mediterranean countries such as France and Italy as well as minor samples from northern Morocco, Algeria, and Tunisia.

The lichen collection has 3300 specimens, which come exclusively from the herbarium of the Pharmacy Faculty (GDA). This collection started in 1980 with materials that came from the research of Dr Manuel Casares and is subsequently enriched with exsiccate exchange with other institutions. Of this collection, 93% comes from Spain, 75% of which were collected in the south-eastern provinces of Granada, Alicante, Almeria, and Jaen. Other well-represented areas include the south-western provinces of Huelva and Cadiz, central provinces of Madrid, Salamanca, and the north-eastern provinces of Tarragona and Zaragoza. We also found a small representation from Morocco (3.3%) and the rest from Austria, Brazil, Georgia, and the Czech Republic.

The fungus collection has been fully computerized and approximately 50% of the lichen collection has also been. Therefore, developing a complete catalogue involved three kinds of type specimens, namely those that: 1) already have a record in the databases; 2) specimens not recorded as a type in the databases but have been identified as such in the collection; and 3) type specimens not recorded in the databases nor identified or treated as such in the collection. For the first kind, quality control was carried out (see next section). For the second kind of type specimens, the literature from the two main researchers and collectors was reviewed. This led to the identification of type specimens that were not included in the databases and were either identified as such in the collection (case 2) or had not been identified as such (case 3). The review of the literature of these authors provided new data to both the collection and the associated databases. All type specimens are now registered in the GDA-GDAC Fungus and Lichen collections and in some cases another number from other herbaria has been included in RelatedInformation or OtherCatalogNumber fields from DarwinCore standard (Wieczorek et al. 2009). This information was provided by gifts and duplicate materials from AH, F, CFB herbaria (Index Herbariorum) and personal herbaria of M. Bon (M.B.) and G. Chevassut (Chev.).

The consistency of data on type-specimen records was verified by comparisons with the information in original publications of the corresponding new species. When some information items such as geographical coordinates, altitude, or identifiers of genetic sequence associated data had not been included in the herbarium database and these data were available in the protologue, they were included in this dataset. The consigned data refer to the original identification for which the nomenclatural type from the taxon name was given. When the nomenclatural type was not recorded as such in the herbarium but was found after searching and reviewing the relevant literature, a revision tag was included in the specimen record. This tag specifies the type status, name, and site where it was validly published and the author who documented it. These data have also been computerized. DARWIN TEST (Ortega-Maqueda and Pando 2008) is the software application used to validate and check records from tables in a Darwincore format before exporting database to a Darwin Core Archive file. DARWIN TEST has been used to check scientific names against the Scientific_Names table from Species 2000 (Species2000 2015), to convert coordinates from UTM to decimal degrees which are used in the Darwincore format and to detect anomalous ASCII characters. Once checked and corrected, these records were exported as a Darwin Core Archive file which was uploaded to the IPT (Integrated Publishing Toolkit) hosted by GBIF.ES (GBIF.ES 2012). The metadata from the dataset have been completed directly in the IPT.

First of all, a query was made to each of the two herbarium databases on nomenclatural types of the fungus and lichen collections. The first database (fungi) is managed with the BIOMEN software application (Delgado et al. 2005), and the second with HERBAR (Pando et al. 1994-2010). Thus the results of the queries were compared with the label information for the type specimens in the Herbarium collection. Three kinds of errors were detected and corrected: first, typographical errors; second, records erroneously described as types; and third, omissions in the database, i.e. types that should have been recorded as such. Publications on the description of new species made by Dr Ortega and Dr Casares, the main collectors and researchers of these collections, were reviewed. After this review, 5 holotypes, 1 isotype, 1 neotype, 1 epitype and 33 paratypes that had not been recorded as types nor had received the appropriate physical curation treatment were detected. After the database update with the inclusion of new types obtained from reviewed publications and data cleaning, the collection corresponding to this dataset was obtained by consulting the database again. Data resulting from this query were manually converted to DarwinCore format. Then, the resulting DarwinCore records were completed by adding the ImageURL and TypeStatus fields, after which it was validated with the DARWIN TEST tool (Ortega-Maqueda and Pando 2008). Finally the DarwinCore Archive was generated to incorporate the metadata in this file and published it on the GBIF.ES Integrated Publishing Toolkit (IPT).

In an effort to fill out the ImageURL field in the step described above, the following steps were taken: after the catalogue of types of fungi and lichens were obtained, all specimens were checked for appropriate physical curation treatment (placed in a type cover, by convention red, to indicate the presence of type specimens and to ensure better protection). Then, digital images of the nomenclatural type and documents included with each specimen were prepared. A Hewlett Packard Scan Jet 5300C model with a resolution of 600 dpi for specimens and 150 dpi for documents was used. After the metadata for each image was included, the corresponding tiled/pyramid TIFF was generated and uploaded at the GBIF.ES “Biodiversity Image Portal of Spanish collections” Hosting and Publishing Service. This collection is available both at “Colección de imágenes de los tipos nomenclaturales de hongos, líquenes, musgos y algas incluidos en el Herbario de la Universidad de Granada (GDA y GDAC)” (http://www.gbif.es/Imagenes.php#GDA-TIPOS-CRIPTO) and also at the Herbarium of University of Granada institutional web (http://herbarium.ugr.es/pages/imagenes/tipos-nomenclaturales/tipos_hongos).

In general, the distribution of the taxa in this catalogue is the western Mediterranean region. Most of the taxa are from the Iberian Peninsula (72 types records from 39 taxa) as shown in Fig. 1, smaller numbers from France (6 types from 5 taxa) and Italy (8 records from 3 taxa) and a single taxon from Morocco. Outside this general geographic distribution, there is a single taxon from Czechoslovakia with 2 records. All these type specimens from outside of Spain were gifts or exchanges from the private herbaria of Dr Ortega’s French and Italian colleagues (G.Chevassut, M.Bon, P.Möenne-Loccoz, and M.Contu). Additionally, type specimens from northern and eastern Spain come from collaboration between Dr Ortega and some researchers such as J. Vila and R. Mahiques from the Societat Catalana de Micologia and Societat Micologica Valenciana, respectively. Most of the records from Spain are from Andalusia (types of 21 taxa from 39 in Spain) and within this region the greatest portion corresponds to the provinces of Granada (10 taxa), Malaga (3), Seville (3), Cadiz (2), Cordoba (1), and Almeria (1), as shown in the Fig. 1.

33°23'60'' and 50°42'0'' Latitude; 6°30'36'' and 15°26'60'' Longitude.

The main taxonomic coverage of this dataset corresponds to Basidiomycota, which constitutes 92% of the type specimens, followed by Ascomycota (6%, included Lecanorales as lichenized fungi) and a minor record of Glomeromycota, as shown in the Fig. 2. Agaricales is the most represented order (91%) and, within this order, specimens of the genus Cortinarius dominate the collection (56%). Fungal diversity in the Mediterranean basin is high. Populations of taxa differing morphologically with respect to their northern vicariants are frequent. Consequently, many taxonomic proposals have been made in order to explain Mediterranean fungal variability, which is common in the case of the genus Cortinarius (Ortega et al. 2008). This is the result of both its overall diversity and the special interest in this genus of the late Dr Antonio Ortega, the main collector and researcher in the fungus collection. As shown in the Fig. 2, Cortinarius is the genus with a greater number of nomenclatural types.

Note: the terms used to complete the type status reference for this dataset are from the International Code of Nomenclature for algae, fungi, and plants (ICN) (McNeill et al. 2012) and Hawksworth (2010) for the obsolete terms which are not included in the current Code. Index Fungorum (2014) has been used to verify the currently accepted names.

Fig. 3 shows the range of the different type status of the catalogue. The highest number of records are paratypes (36) followed by holotypes (25), isotypes (18), clastotypes (5 fragments from holotypes, one from neotype), and a single neotype, epitype and topotype.

Of the 49 taxa, 34 are accepted in their original position (marked by an asterisk* in the taxonomic ranks) and four had already been combined into other genera: Sarcodon mediterraneus A.Ortega & Contu to Beenakia, Rugosomyces pudicus Bon & Contu to Calocybe, Glomus custos C. Cano & Dalpé to Rhizophagus, and Bacidia subtilis Vêzda to Fellhanera. One proved to be a synonym of a previously described species (Cortinarius haasii var. quercus-ilicicola A.Ortega, Suár.-Sant. & J.D.Reyes of Cortinarius callochrous (Pers.) Gray), another has been combined into a supraspecific rank (Gymnopus dryophilus var. lanipes (Malençon & Bertault) A.Ortega, Antonín & Esteve-Rav.), and the rest (9) have not been accepted into the proposed infraspecific rank (variety and one form): Conocybe arrhenii var. squamosipes A.Ortega & Esteve-Rav., Cortinarius assiduus var. plesiocistus A.Ortega, Vila & Bidaud, Cortinarius caesiostramineus var. cadinanos-aguirrei Moënne-Locc. & A.Ortega, Cortinarius caerulescens var. praetermissus (Bergeron ex Reumaux) A.Ortega & Moënne-Locc., Cortinarius vernus var. nevadavernus Suár.-Sant. & A.Ortega, Entoloma griseocyaneum var. glyciosmus Esteve-Rav. & A.Ortega, Hydropus floccipes var. luteipes A.Ortega & M.Zea, Marasmiellus virgatocutis var. parvisporus Esteve-Rav. & A.Ortega, Baeospora myosura f. xeruloides A.Ortega & Esteve-Rav.

Fungi

9b8b659f-8470-4445-926c-8b3f2bc32415

GDA-GDAC

Dried

Curatorial unit: 88 with an uncertainty of 0 (specimens); curatorial unit: 49 with an uncertainty of 0 (species); curatorial unit: 19 with an uncertainty of 0 (genera); curatorial unit: 6 with an uncertainty of 0 (order)

Darwin Core Archive Catalogue of type specimens of fungi and lichens deposited in the Herbarium of the University of Granada (Spain)

http://www.gbif.es:8080/ipt/archive.do?r=gda-fungi-tipos

7ac0504d-0230-4029-afbe-04657ae47c48

1