Purpose: Existing knowledge on the distribution of harvestmen throughout the Iberian Peninsula is still highly fragmented (Prieto 2003). Several studies on particular genera (e.g., Prieto 2004, Prieto and Fernández 2007, Merino-Sáinz et al. 2013a), as well as studies with planned, repeated, and systematic samplings in some locations (Merino-Sáinz and Anadón 2008, 2013, Rosa García et al. 2009a, b, 2010a, b, Merino-Sáinz et al. 2013b) have contributed to improving this knowledge. Global or specific studies on biodiversity are also enabled by the review, digitisation, and data release of specimens housed in biodiversity collections at research centers, universities, museums, and in the possession of individuals. These practices facilitate the identification of gaps in our knowledge of taxa distribution across space and time.

Within this context, biodiversity data on specimens from the BOS Arthropod Collection (hosted at the Department of Organisms and Systems Biology, (Spanish acronym BOS), University of Oviedo) are being digitised and the data released through the Global Biodiversity Information Facility (GBIF) data-portal (Department information as data published and available datasets: http://www.gbif.org/publisher/95cb537c-74c5-4c1e-ae24-32e7ea08f380; general digitisation and data release workflow of the BOS Arthropod Collection: Torralba-Burrial and Ocharan 2013). However, there is a need to establish priorities in the digitisation of specimens data of biodiversity collections (see Berents et al. 2010 for different approaches), especially in situations where mass digitisation methods are not available (see Beaman and Cellinese 2012). As such, we evaluate whether the effort of reviewing and digitising (harvestmen) specimens from unplanned collection events can provide useful data on their biodiversity and distribution, or whether it is better to limit digitisation to only those specimens associated with standardised samplings (planned collection events), which provide quantitative data in each location and allow for comparisons between locations over time.

The BOS Arthropod Collection includes harvestmen from the northern part of the Iberian Peninsula that, since 1977, have been obtained through systematic repeated sampling in several locations, as well as through non-harvestmen-specific sampling and accidental occurrences. Specimens were collected systematically from the Muniellos Biosphere Reserve between 2000 and 2002 (Merino-Sáinz and Anadón 2008), and from several locations in the provinces of Asturias, Cantabria, and Pontevedra between 2009 and 2011 (Merino-Sáinz et al. 2013c describe the harvestmen subcollection of BOS). Therefore, we decided to study and compare the data derived from unplanned collections events (untargeted sampling) with these data derived from planned, standardised, and periodic sampling. We have combined these analyses with the published results of similar studies using pitfall traps in western Asturias (Rosa García et al. 2009a, b, 2010a, b). In effect, this is the first data-paper to employ a hybrid approach, wherein the subset metadata from a large published dataset are described and a comparative analysis is carried out, in order to evaluate digitisation priorities. The aims of this paper are, thus, to 1) test whether the effort of reviewing and digitising (harvestmen) specimens from unplanned collection events can provide useful data about their distribution and biology, and 2) assess possible biases arising from the use of this type of data.

Project title: Informatización de la Colección de Artrópodos BOS de la Universidad de Oviedo / Digitisation of the BOS Arthropod Collection of University of Oviedo

Personnel digitisation and metadata creator: A. Torralba-Burrial

Administrative contact: A. Anadón

BOS-Opi determination specialist: I. Merino-Sáinz

Subset collectors: Most of the collectors provided less than ten harvestmen records in this subset. Only Merino-Sáinz collected more than 10 specimens. All of the collectors are listed in Supplementary material 1 (http://hdl.handle.net/10651/24734) next to each specimen.

Funding: The digitisation of this biological collection was supported by the Spanish National R+D+i Plan (MICINN, Spanish Government, grant ref. PTA2010-4108-I) and PCTI Asturias (Asturias Regional Government, ref. COF11-38) through a contract with ATB.

Specimens were identified by IMS, which was supported by a Severo Ochoa pre-doctoral grant (ref. BP08039, FICYC, Asturias Regional Government).

Study area descriptions/descriptor: Harvestmen data in the subset are from the same area as the full Opiliones of the BOS Arthropod Collection dataset. Specimens are mainly from the northern third of the Iberian Peninsula (chiefly the Spanish provinces of Asturias, Cantabria, and León, with a few records from other neighbouring provinces) (see Figure 1).

Data sources of harvestmen data from planned collection events with pitfall trapping: Merino-Sáinz and Anadón (2008, 2013); Merino-Sáinz et al. 2013c, Rosa García et al. (2009a, b, 2010a, b).

Design description: The data subset is part of the large dataset of Opiliones housed in the BOS Arthropod Collection (Universidad de Oviedo 2013-). Details of the digitisation process are available in the description of the BOS Collection digitisation workflow (Torralba-Burrial and Ocharan 2013) and in the data paper on the harvestmen subcollection (Merino-Sáinz et al. 2013c). In that data-paper, we argue that the large dataset could be used to assess, among other things, the importance of unplanned collection data in filling in knowledge gaps if planned (standardised sampling) collection data are not available or are incomplete. With this aim in mind, we chose a data subset from the harvestmen subcollection, which included data only from unplanned collection events. This subset was used to compare specimens data with published data obtained through planned, standardised, and periodic samplings using pitfall traps in several locations in the north of the Iberian Peninsula (see Merino-Sáinz and Anadón 2013 for a checklist of the species found in the studies). Moreover, we used all of the published data on Iberian harvestmen, not just the BOS Arthropod Collection harvestmen data, to analyse the distributional knowledge gained by digitising this subset, e.g., the first provincial records. Figure 2 shows a diagram depicting the methodological design of our analysis.

General taxonomic coverage description: Seventeen taxa were identified to the species level. Due to the biological phase or sex of the specimens, or unresolved taxonomic issues, 39 records (8%) were assigned only to the genus level. Those specimens belonging to the genus Paramiopsalis represent species number 18. The numbers of records per species and per family (also including specimens identified to the genus level) are shown in Table 1.

The family Phalangidae comprised the largest number of identified species (seven), followed by Sclerosomatidae (five). However, when the number of records is considered, Sclerosomatidae was the most frequent family (around one hundred records for both Leiobunum blackwalli and Leiobunum rotundum), followed by Phalangidae, with only one species Phalangium opilio with a high number of records, similar to the Leiobunum species, and other species with only a few records. Five families (and the remaining species) had less than 30 records each.

Kingdom: Animalia

Phylum: Arthropoda

Class: Arachnida

Order: Opiliones

Family: Sclerosomatidae, Phalangiidae, Ischyropsalididae, Nemastomatidae, Trogulidae, Sironidae

Common names: Animals, Arthropods, Arachnids, Harvestmen

1977–2012.

Parent collection identifier: Colección de Artrópodos BOS

Collection name: Colección de Artrópodos BOS de la Universidad de Oviedo: Opiliones (BOS-Opi) subset recolecciones no planeadas

Collection identifier: http://www.gbif.org/dataset/7cebf715-c3b0-4477-99e7-f6f3aca27bbe

Curatorial unit: 472 with an uncertainty of 0 (Data records)

Curatorial unit: 536 with an uncertainty of 0 (Specimens)

Method step description: This data subset was extracted from the large dataset of harvestmen in the BOS Arthropod Collection (University of Oviedo 2013-). Specimens data in the subset are listed in Supplementary material 1 - Appendix A (http://hdl.handle.net/10651/24734), which includes the municipality, location, date, sampling method, amount, sex, and collector of the 536 taxonomically identified specimens corresponding to BOS-Opi codes 493-960. Using these codes, most of the specimen data (including their georeferenced locations) are available in reusable format in the DarwinCore Archive of the data-paper describing the BOS-Opi subcollection (Merino-Sáinz et al. 2013c) and through the GBIF data-portal (Universidad de Oviedo 2013-, http://data.gbif.org/datasets/resource/15038).

Bibliographic records on each harvestmen taxon (except Odiellus spinosus) are listed in Merino-Sáinz and Anadón (2008, 2013) and Merino-Sáinz et al. (2013a). In the faunistic analysis, each species was considered in accordance with its general distribution as Iberian endemic, European element, or Holarctic element (see Merino-Sáinz and Anadón 2008). Specimens identified as Trogulus nepaeformis belong to a related undescribed species probably endemic to the Iberian Peninsula, according to Schönhofer and Martens (2010). However, distribution data on this undescribed species are not available; thus, the chorological type (European element) is retained to compare data with previous articles (Merino-Sáinz and Anadón 2008, 2013) and to test whether biases exist.

We conducted a hierarchical cluster analysis (group average clustering algorithm, see algorithm choice discussion in Clarke and Gorley 2006) on similarity matrices, in order to compare this subset with the data obtained through planned, standardised, and periodic samplings using pitfall traps in several locations in the north of the Iberian Peninsula (Merino-Sáinz and Anadón 2013). The inventory of each locality included all of the harvestmen species sampled using the pitfall traps in that locality; subset inventory included all of the harvestmen species present in the unplanned events collection. Only qualitative data on species presence, rather than abundance data, were used in the analysis, and the similarity matrices were calculated using a species-presence Sørensen index (Sørensen 1948) (only positive results, i.e., species present in a pair of inventories, incremental similarity between inventories, species absent from both inventories -double negative- don’t). Data were not standardised through sampling efforts, because the aims of the analysis were to compare the results of standardised sampling data with unplanned sampling data employing very different sampling and identification efforts. The analysis was carried out using the PRIMER V6 software (Clarke and Gorley 2006).

Study extent description: Harvestmen specimens included in the subset came from different localities in the Iberian Peninsula, at different distances from one another, and were collected at different dates between 1977 and 2011. Nonetheless, most of them came from the north of the Iberian Peninsula, and all of them came from the northern half (see Figure 1). The heterogeneity of the localities, most of which are only represented by a single sample or even only an isolated specimen, means that a general list of localities is not useful to short data exposition; rather, the locations are listed beside each specimen in Supplementary material 1 - Appendix A (supplementary file http://hdl.handle.net/10651/24734).

Harvestmen data obtained through planned collection using pitfall traps and deposited in the BOS Arthropod Collection included specimens collected from the Muniellos Biosphere Reserve between 2000 and 2002 (Merino-Sáinz and Anadón 2008) and from several locations in the provinces of Asturias (Muros, Oviedo, Villar), Cantabria (Vioño) and Pontevedra (Panjón) collected between 2009 and 2011 (Merino-Sáinz and Anadón 2013, Merino-Sáinz et al. 2013c). We also referred to published results of standardised pitfall samplings in western Asturias (Illano: Rosa García et al. 2009a, b, 2010a, b) for comparison purposes, as they were collected from the same areas (see Figure 1).

Sampling description: We studied a data subset of the harvestmen specimens in the BOS Arthropod Collection at the University of Oviedo that had been directly collected (by hand) on entomological field trips by students and lecturers from this department (listed in Supplementary material 1: http://hdl.handle.net/10651/24734, beside each specimen). This subset also included our own data obtained using diverse methods—collecting directly by hand, beating vegetation over an upturned umbrella, and using Berlese funnels, light traps, Malaise traps, and sieves; only 6% of collections used pitfall traps. Therefore, the specimens included in this study did not derive from harvestmen-targeted research projects, theses, or historical collections, but were collected at random with no prior sampling design.

Metadata language: English

Date of metadata creation: 2014-02-05

Hierarchy level: Subset

Metadata distribution: http://www.gbif.es:8080/ipt/resource.do?r=bos-opi_unplanned_collection_events

Format name metadata: Ecological Metadata Language (EML) and HTML in web.

Data distribution: BOS-Opi dataset http://www.gbif.es:8080/ipt/archive.do?r=Bos-Opi

Subset codes in dataset: BOS-Opi codes 493-960.

Publication date of data: 2013-07-04

Update police: Subset will not be updated.

Licences of use: This BOS Arthropod Collection of University of Oviedo (Spain): Opiliones unplanned collection events subset, as part of BOS Arthropod Collection Dataset: Opiliones (BOS-Opi) dataset is made available under the Open Data Commons Attribution License: http://www.opendatacommons.org/licenses/by/1.0/.

A small subcollection of harvestmen from the north of the Iberian Peninsula, gathered using non-Opiliones-targeted sampling methods and in many cases by non-specialist collectors, presented a high species richness similar to planned, periodic, and costlier studies. This subcollection enabled us to extend our knowledge on the distribution of 18 species. The 536 specimens in the subset showed very interesting faunistic results, while less effort was exerted on identification and digitisation than in planned, periodic collection events using pitfall traps. The data subset contained six first provincial records of various species; Phalangium opilio locations in eight provinces without previous data were also provided. Nevertheless, we also detected some drawbacks to this type of data collection; collection was biased towards adults of larger species (with long legs or wide bodies) occupying shrubby or arboreal habitats, which may also affect the biogeographic analysis of the dataset. Nevertheless, this study highlights the importance of the general biodiversity collections in museums and at universities and the need to digitise their specimens, including the data from non-targeted, or unplanned, samplings, especially when poorly studied groups are involved. The digitisation of unplanned collections can help to justify the investments made to collect, house, and study such collections. Moreover, it is important to keep in mind that most of the collections at the university/museum, NGO, and amateur scientist levels are not comprised of data collected through planned events, but mainly through unplanned events. The digitisation of such unplanned collections has great potential to (1) bridge gaps in existing knowledge, and (2) strengthen existing understanding about the status of biodiversity.