Dataset of MIGRAME Project (Global Change, Altitudinal Range Shift and Colonization of Degraded Habitats in Mediterranean Mountains)

Abstract In this data paper, we describe the dataset of the Global Change, Altitudinal Range Shift and Colonization of Degraded Habitats in Mediterranean Mountains (MIGRAME) project, which aims to assess the capacity of altitudinal migration and colonization of marginal habitats by Quercus pyrenaica Willd. forests in Sierra Nevada (southern Spain) considering two global-change drivers: temperature increase and land-use changes. The dataset includes information of the forest structure (diameter size, tree height, and abundance) of the Quercus pyrenaica ecosystem in Sierra Nevada obtained from 199 transects sampled at the treeline ecotone, mature forest, and marginal habitats (abandoned cropland and pine plantations). A total of 3839 occurrence records were collected and 5751 measurements recorded. The dataset is included in the Sierra Nevada Global-Change Observatory (OBSNEV), a long-term research project designed to compile socio-ecological information on the major ecosystem types in order to identify the impacts of global change in this mountain range.


Rationale
Currently, there is strong scientifi c evidence of the eff ects of global change on natural systems (Parmesan 2006, Rosenzweig et al. 2008, García et al. 2014, O'Connor et al. 2015. Some ecological processes are being altered due to the changing climate, such as species distribution (Th uiller et al. 2005, Lenoir et al. 2008, phenology (Parmesan and Yohe 2003, Gordo and Sanz 2010, Wolkovich et al. 2014, ecological interactions (Hughes 2000, Suttle et al. 2007); among others. Land-use changes and climate change are the most important drivers of biodiversity shifts (Sala et al. 2000).
One of the most obvious biotic responses from global warming are the latitudinal and altitudinal shifts of species and communities (Allen and Breshears 1998, Jump and Peñuelas 2005, Lenoir et al. 2008. Species tend to expand into new areas that are becoming favourable, and retract from those that turn hostile. In consideration of two main drivers of global change (climatic warming and land abandonment), an understanding of the dynamics of altitudinal migration and colonization of marginal habitats is critical in order to develop eff ective forest-management strategies.
Th e project Global Change, altitudinal range shift, and colonization of degraded habitats in Mediterranean mountains (MIGRAME) was designed to assess the capacity of altitudinal migration and colonization of marginal habitats by a Mediterranean forest ecosystem . Th is assessment considers two global change drivers: temperature increase and land-use changes. In so doing, this project analyzes the pattern of altitudinal migration and colonization of marginal habitats by a vulnerable ecosystem in a Mediterranean mountain region, which represents the rear edge of their distribution: forests of Quercus pyrenaica Willd.
Th e Mediterranean region has shown broad climate shifts in the past (Luterbacher et al. 2006) and is potentially vulnerable to forthcoming climatic changes (Pacifi ci et al. 2015), being considered a key region in future climate-change projections (Giorgi 2006, Giorgi andLionello 2008). Concomitantly, land-use changes are considered a major driver of vegetation change (McGill 2015). Th is is especially relevant in Mediterranean region, which has undergone intense antrophic activities for millennia (Padilla et al. 2010) shaping the current landscape (Valbuena-Carabaña et al. 2010).
In this context, Mediterranean ecosystems are considered natural laboratories in which to study global change, due to their high sensitivity to global-change drivers (Matesanz andValladares 2014, Doblas-Miranda et al. 2015).

Study area descriptions/descriptor
Th e target ecosystem of the project encompasses the Pyrenean oak forests (Quercus pyrenaica Willd.) of Sierra Nevada.
Sierra Nevada is a high-mountain range located in southern Spain (37°N, 3°W) with altitudes of between 860 m and 3482 m a.s.l. Th e climate is Mediterranean, characterized by cold winters and hot summers, with pronounced summer drought (July-August). Th e Sierra Nevada mountain range hosts a high number of endemic plant species (c. 80) (Lorite et al. 2007) for a total of 2,100 species of vascular plants (25% and 20% of Spanish and European fl ora, respectively), and thus it is considered one of the most important biodiversity hotspots in the Mediterranean region (Blanca et al. 1998). Th is mountain area has 27 habitat types (listed in the European Union Habitat Directive) harbouring 31 animal species (20 birds, 5 mammals, 4 invertebrates, 2 amphibians and reptiles) and 20 plant species listed in the Annex I and II of EU Habitat and Bird Directives. Sierra Nevada has several types of legal protection: Biosphere Reserve MAB Committee UNESCO; Special Protection Area and Site of Community Importance (Natura 2000 network); and National Park. Th ere are 61 municipalities with more than 90,000 inhabitants. Th e main economic activities are agriculture, tourism, beekeeping, mining, and skiing (Bonet et al. 2010).
For a description of the Pyrenean oak forests in Sierra Nevada, see Study extent description section.

Design description
Th e specifi c aims of the MIGRAME project are: • To analyse the relevance of altitudinal migration at the leading edge (high elevation) of the range distribution of Pyrenean oak formation. • To analyse the importance of the recolonization process of marginal habitats (abandoned croplands and pine plantations) close to Pyrenean oak formation. Derived from the two global-change drivers, we have considered two main hypothesis ( Figure 1):

Altitudinal migration hypothesis
Several studies have pointed out a trend towards higher temperatures and lower precipitation for the Mediterranean area (Giorgi andLionello 2008, García-Ruiz et al. 2011). Climate projections forecast an increase of +4.8 °C at the end of the 21st century (Benito et al. 2011) for Sierra Nevada. In this context, shifts in the altitudinal (and latitudinal) distribution of species and communities are expected (Th uiller et al. 2008, Gottfried et al. 2012. We hypothesised that the range shift of Q. pyrenaica in Sierra Nevada is changing as a consequence of recent changes to temperature, and we would expect an upward expansion ( Figure 1a).

Marginal habitat colonization hypothesis
In Mediterranean area, cropland abandonment has been widespread during the second half of the last century (Valbuena-Carabaña et al. 2010, Pías et al. 2014). Land-use change models predict an increase in this trend in the future (Rounsevell et al. 2006). In fact, land abandonment is considered one of the most powerful global-change drivers in developed countries (Escribano-Avila et al. 2012).
We hypothesised that the land-use changes in high mountain (e.g. abandonment of croplands, management of pine plantations) should facilitate the native forest regeneration, and a process of colonization of marginal habitat (abandoned cropland, pine plantations) will occur ( Figure 1b). Overall, focusing on changes will occur in altitudinal migration and/or colonization of marginal habitats, we examine the following questions: Are altitudinal changes in Pyrenean oak forests associated with recent climate changes? Are they more consistent with changes in land use, or are they consistent with both global-change drivers?

Taxonomic coverage
Th is dataset includes records of the phylum Magnoliophyta (3823 records, 99.58%) and marginally Pinophyta (16 records, below 1% of total records). Most of the records included in this dataset belong to the class Magnoliopsida (99.58%). Th ere are 5 orders represented in the dataset, with Fagales (98.98%) being the most important order. Th e other 4 orders (Rosales, Cupressales, Sapindales and Pinales) represent only 1.02% of the records. In this collection, 5 families are represented: Fagaceae, Rosaceae, Cupressaceae, Pinaceae, and Sapindaceae. Th e most represented taxa are Quercus pyrenaica Willd. and Quercus ilex L. (81.74 and 17.24%, respectively). Of the six taxa included on the dataset, three are considered threatened (Table 1).

Taxonomic ranks
Kingdom: Plantae Phylum: Magnoliophyta, Pinophyta Class: Magnoliopsida (Dicotyledones), Pinopsida  Th is species reaches its southernmost European limit at Sierra Nevada mountains, where eight oak patches (2400 Has) have been identifi ed (Figure 2b), ranging between 1100 and 2000 m a.s.l. and generally associated to major river valleys. Sierra Nevada is considered a glacial refugia for deciduous Quercus species during glaciation (Brewer et al. 2002, Olalde et al. 2002, Rodríguez-Sánchez et al. 2010) and these populations are considered as a rear edge of the habitat distribution, which is important in determining habitat responses to expected climate change (Hampe and Petit 2005).
Th ese forests, like other vegetation types, have undergone intense human pressure (wood cutting, grazing, etc.) which has reduced their distribution area and in some cases has altered their fl oristic pattern (Gavilán et al. 2000, Gavilán et al. 2007. Q. pyrenaica is considered as vulnerable in southern Spain . Th e populations of Pyrenean oak forests at Sierra Nevada are considered relict forests (Melendo andValle 2000, Vivero et al. 2000), undergoing intensive anthropic use in the last few decades (Camacho-Olmedo et al. 2002, Valbuena-Carabaña et al. 2010. Th e relict presence of this species in Sierra Nevada is related both to its genetic resil- ience as well as to its high intraspecifi c genetic diversity (Valbuena-Carabaña and Gil 2013). However, they are also expected to suff er the impact of climate change, due to their climate requirements (wet summers). Th us, simulations of the climate change effects on this habitat forecast a reduction in suitable habitats for Sierra Nevada (Benito et al. 2011).

Dataset of MIGRAME Project (Global Change, Altitudinal Range Shift and Colonization of Degraded Habitats in Mediterranean Mountains)
Collection identifier http://www.gbif.es/ipt/resource.do?r=migrame

Study extent description
Th e MIGRAME dataset covers the Pyrenean oak forests (see Spatial coverage section) in Sierra Nevada mountain range (see Study area descriptions section).

Sampling description
We sampled two localities of the Pyrenean oak forests in Sierra Nevada: Robledal de Cañar and Robledal de San Juan. We selected those two sites based on previous works (Pérez-Luque 2011) that clustered the populations of Q. pyrenaica forests based on their plant species composition and environmental features. Th e Robledal de Cañar site (Figure 2c) (1366-1935 m a.s.l., 37°57'28.04"N, 3°25'57.1"W; Cáñar, Granada, SE Spain) was located in the Alpujarras Region on the southern slopes of Sierra Nevada. Th e Robledal de San Juan (Figure 2d) (1189-1899 m a.s.l., 37°7'29.63"N, 3°21'54.60"W; Güejar-Sierra, Granada, SE Spain) site was located in the northern slopes of Sierra Nevada.
Th e sampling design was determined by the hypothesis of the project (see Project Design description section).

Altitudinal migration design
To test our hypothesis of altitudinal migration, we sampled a total of 104 transects ( Table 2) distributed along an altitudinal gradient at the two sites. We sampled two transects (at least 10 m apart) every 25 m of elevation from forest limit to treeline ecotone at both study sites. At each locality, we performed three replicates of this design (Figure 3a).

Habitat colonization design
To test the hypothesis of colonization of marginal habitats, we laid out transects in two types of marginal habitats: abandoned agricultural areas and pine plantations ( Figure  3b). A total of 64 transects were located within the marginal habitat and on the edge between marginal habitat and Pyrenean oak forest. Th e number of transects inside the marginal habitat was determined by the size of the marginal habitat (Table 3).

Forest samplings
In addition to the above surveys, we conducted a survey inside Q. pyrenaica forests. A total of 31 transects were distributed at the two sites.

Data collection
We sampled a total of 199 linear transects of 30 m × 10 m (Suppl. material 1). Within each transect, all tree species were recorded and the species identity was recorded. Diameter size and tree height were measured for all individuals. Field data were recorded using handheld PDAs. A customized application (app) (Figure 4) was built to facilitate both data collection and storage (Pérez-Pérez et al. 2013 -http://obsnev.es/noticia. html?id=4513). Th e data were automatically integrated into an information system using this application.

Method step description
All data were stored in a relational database (PostgreSQL) and added to the Information System of Sierra Nevada Global-Change Observatory (Figure 4) (http://obsnev. es/linaria.html -Pérez-Pérez et al. 2012; Free access upon registration). Taxonomic and spatial validations were made on this database (see Quality-control description). A custom-made SQL view of the database was performed to gather occurrence data and other variables associated with some occurrence data (diameter size and tree height of each individual). Th e occurrence and measurement data were accommodated to fulfi l the Darwin Core Standard (Wieczorek et al. 2009, Wieczorek et al. 2012. We used Darwin Core Archive Validator tool (http://tools.gbif.org/dwca-validator/) to check whether the dataset met Darwin Core specifi cations. Th e Integrated Publishing Toolkit (IPT v2.0.5) (Robertson et al. 2014) of the Spanish node of the Global Biodiversity Information Facility (GBIF) (http://www.gbif.es/ipt) was used both to upload the Darwin Core Archive and to fi ll out the metadata.

Quality control description
Transects coordinates were recorded with a handheld Garmin eTrex Vista Global Positioning System (GPS, ±5 m accuracy, Garmin (2007)) (WGS84 Datum). We also used colour digital orthophotographs provided by the Andalusian Cartography Institute and GIS (ArcGIS 9.2; ESRI, Redlands, California, USA) to verify the geographical coordinates of each sampling plot (Chapman and Wieczorek 2006). Th e specimens were taxonomically identifi ed using Flora iberica (Castroviejo 1986(Castroviejo -2005. Th e scientifi c names were checked with databases of International Plant Names Index (IPNI 2013) and Catalogue of Life/Species 2000 (Roskov et al. 2015). We also used the R package taxize Szöcs 2013, Chamberlain et al. 2014) to verify the taxonomical classifi cation.