Biovera-Epi: A new database on species diversity, community composition, and leaf functional traits of vascular epiphytes along an elevational gradient in Mexico

This data paper describes a new, comprehensive database (BIOVERA-Epi) on species distributions and leaf functional traits of vascular epiphytes, a poorly studied plant group, along gradients of elevation and forest-use intensity in the central part of Veracruz State, Mexico. The distribution data includes frequencies of 271 vascular epiphyte species belonging to 92 genera and 23 families across 120 20 m × 20 m forest plots at eight study sites along an elevational gradient from sea level to 3500 m a.s.l. In addition, BIOVERA-Epi provides information on 1595 measurements of nine morphological and chemical leaf traits from 474 individuals and 102 species. For morphological leaf traits, we provide data on each sampled leaf. For chemical leaf traits, we provide data at the species level per site and land-use type. We also provide complementary information for each of the sampled plots and host trees. BIOVERA-Epi contributes to an emerging body of synthetic epiphytes studies combining functional traits and community composition.
 BIOVERA-Epi includes data on species frequency and leaf traits from 120 forest plots distributed along an elevational gradient including six different forest types and three levels of forest-use intensity. It will expand the breadth of studies on epiphyte diversity, conservation, and functional plant ecology in the Neotropics and will contribute to future synthetic studies on the ecology and diversity of tropical epiphyte assemblages.


Introduction
Elevational gradients provide a wide range of opportunities for studying the effects of different ecological and evolutionary factors on biodiversity patterns. Steep elevational gradients in temperature, precipitation, and other climatic variables usually play a fundamental role in shaping plant diversity Grytnes 2010, Peters et al. 2019), and also contribute to linkages between plant traits and environmental conditions ( Bruelheide et al. 2018, Keddy 1992. They are also used as proxies for understanding diversity patterns across latitudinal gradients (McCain and Grytnes 2010), while controlling for species pools and biogeographic history (Ricklefs 2004). Additionally, anthropogenic forest disturbance may modify climatic conditions at local and regional scales, which in turn may affect the response of species, especially for canopy-dwelling life forms such as vascular epiphytes that are sensitive to changes in air humidity and temperature (Larrea and Werner 2010, Werner and Gradstein 2009, Zotz and Bader 2009.
Functional traits are measurable characteristics of individual plants impacting their growth, reproduction and survival (Violle et al. 2007) and reflect how species interact with their environment (Vesk 2013). Functional traits are widely used to elucidate mechanisms that underpin many ecological processes along vertical and horizontal environmental gradients (e.g. Petter et al. 2015, Bruelheide et al. 2018) but also evolutionary patterns associated with variation in plant form and function, such as geographic distributions of woody and non-woody species . Despite recent progress (e.g. Agudelo et al. 2019, Petter et al. 2015, studies in the field of functional traits of vascular epiphytes are rare, suggesting that our knowledge of the factors that determine the distribution of vascular epiphytes along environmental gradients is similarly limited. Deforestation and forest fragmentation represent major threats to biodiversity, as well as to ecosystem integrity and functioning (Tapia-Armijos et al. 2015). Furthermore, increasing temperatures and changing precipitation patterns may negatively affect mountain biodiversity, causing upward shifts in the treeline (Cazzolla Gatti et al. 2019), and shifting the distribution of plants and animals (McCain et al. 2016). While a growing number of studies shows that climate change affects a wide range of species and ecosystems (Peters et al. 2019, Root et al. 2003, Trisos et al. 2020, Walther et al. 2002, the effects of deforestation and fragmentation on tropical mountain ecosystems are still poorly understood (Payne et al. 2017). Due to their dependence of trees, vascular epiphytes are particularly vulnerable to these changes (Barthlott et al. 2001, Köster et al. 2009).
Mexico is a country with high floristic diversity and endemism. Almost 50% of its 23,114 native species of vascular plants are endemic. Thus, Mexico ranks fourth in species richness globally, after Brazil, China, and Colombia, and is second in terms of endemism ( Villaseñor and Ortiz 2014). However, Mexico has lost approximately half of its forest cover in the past 50 years (Barsimantov and Kendall 2012). Although deforestation rates have been declining in recent years, the country lost 155,000 ha/year between 2000 and 2005 (Barsimantov and Kendall 2012, Food and Agriculture Organization 2010, Velzquez et al. 2002. The Mexican state of Veracruz, has one of the highest rates of deforestation with more than 80% of primary vegetation having been converted to pastures, plantations, and secondary vegetation (Ellis et al. 2011, Williams-Linera et al. 2002. Given its species richness and endemism (c. 30% of 8500 vascular plant species are endemic to Mexico; Villaseñor and Ortiz 2014), Veracruz also plays an important role in biodiversity conservation (Gómez-Pompa et al. 2010, Sarukhán et al. 2014. It has been estimated that about 7.8% of the Mexican vascular flora are epiphytes, 750 of which (569 angiosperms and 181 pteridophytes) are native to Veracruz (Krömer et al. 2020). Vascular epiphytes usually reach their highest diversity in humid tropical forests at mid elevations ( Guzmán-Jacob et al. 2019, Küper et al. 2004, Körmer et al. 2005, Cardelús et al. 2006. Moreover, they contribute significantly to ecosystem functioning through biotic interactions and by providing microhabitats for other organisms (Nadkarni 1984, Veneklaas et al. 1990, Zotz 2016). Our study sites in the central part of Veracruz, host a wide variety of different ecosystems including tropical semi-humid deciduous forest and humid montane and pineoak forests (Williams-Linera et al. 2007, Carvajal-Hernandez et al. 2020 and have a diverse epiphyte flora (Krömer et al. 2020).

General description
Purpose: BIOVERA-Epi includes plot data from an elevational gradient located in the central part of the State of Veracruz, Mexico. Specifically, it contains two distinct but related datasets: the first dataset includes distribution and plot level frequency information (frequency.subplot)for 271 vascular epiphyte species, sampled in 120 20 m × 20 m plots along the elevational gradient, ranging from 0 to 3500 m a.s.l. The second dataset includes measurements of nine morphological and chemical leaf traits for 102 species, 474 individuals and a total of 1595 leaves, which were sampled in 45 plots at three sites along the same elevational gradient. The leaf traits studied were: leaf area, leaf density, specific leaf area (SLA), leaf dry matter content (LDMC), leaf nitrogen content, leaf phosphorus content, leaf carbon content, nitrogen isotope ratio (d N), and carbon isotope ratio (d C). For each plot, we also provide geographical coordinates, forest-use intensity (old-growth, degraded, secondary), and elevation. For the surveyed host trees, we report diameter at breast height (DBH), total height (H), and species identity (see Data collection).
The elevational gradient spanned from sea level to 3500 m on the eastern slopes of Cofre de Perote, a 4282 m extinct volcano located in the central part of Veracruz State, Mexico (Fig. 1). In this region, the Trans-Mexican volcanic belt and the Sierra Madre Oriental converge, creating complex geological conditions and combining floristic elements from the Nearctic and Neotropics. The climate in the study region ranges from dry and hot in the lowlands (mean annual temperature (MAT): 25 °C; mean annual precipitation (MAP): 1222 mm yr ) to humid and temperate at mid-elevations (MAT: 13-19 °C; MAP: 2952-1435 mm yr ) and dry and cold at high elevations (MAT: 9 °C; MAP: 708 mm yr ; data according to the National Meteorological Service of Mexico 1951Mexico -2010. Along the elevational gradient, six main vegetation types are commonly found (Carvajal-Hernández and Krömer 2015): (1) semi-humid deciduous forest at 0-700 m, (2) tropical oak forest at 700-1300 m, (3) humid montane forest at 1300-2400 m, (4) pine-oak forest at 2400-2800 m, (5) pine forest at 2800-3500 m and (6) fir forest at 3500-3600 m.
We investigated three levels of forest-use intensity (FUI) that could consistently be found along the entire gradient (following Gómez- Díaz et al. 2017): (1) old-growth forests (OG) encompass mature forests with no or only few signs of logging and other human impacts, and are classified as the lowest FUI; (2) degraded forests (DF) are forests with clear signs of past logging, sometimes with ongoing cattle grazing, removal of understory and/or harvesting of non-timber forest products, and are classified as intermediate FUI; and (3) secondary forests (SF) represent forests at an intermediate successional stage 15-25 years after abandonment (based on interviews with local landowners), often with signs of continued human impacts, such as the removal of understory vegetation, non-timber forest products or partial tree cutting and occasional cattle grazing, and are classified as high FUI.

Data collection: species distribution
We selected eight study sites each separated by c. 500 m in altitude along the elevational gradient representing the following elevational ranges: 0-45 m, 610-675 m, 980-1050 m, 1470-1700 m, 2020-2200 m, 2470-2600 m, 3070-3160 m, and 3480-3545 m. At each study site, we surveyed vascular epiphytes in five non-permanent 20 m × 20 m plots for each of the three FUI levels respectively yielding a total of 120 plots (Suppl. material 1). We used a Garmin® GPSMAP 60Cx device (Garmin International, Inc. Kansas, USA) to record geographical coordinates and elevation of all plots. . We selected one mature host tree per plot based on size, vigor, and crown structure for safe canopy access. We climbed from the base to the outer portion of the tree crown using the single-rope climbing technique (Perry 1978) and recorded the presence of vascular epiphyte species in each of the five vertical tree zones according to Johansson (1974), (Fig. 2). Johansson zones are a frequently used scheme to record and describe the spatial -1 -1 -1 distribution of vascular epiphytes within tree trunks and canopies (Gradstein et al. 2003, Sanger andKirkpatrick 2016). We recorded diameter at the breast height (DBH) and total height for each climbed tree. We recorded the frequency of each species as the sum of incidences in the four nested subplots and the central host tree (frequency.subplot, maximum frequency per plot = 4) (Suppl. material 2, Figs 3, 4). We also recorded the frequency of each species as the sum of incidences in the five Johansson zones of the central host tree (frequency.J.zones, maximum frequency = 5).

Data collection leaf trait dataset
In a separate sampling campaign from June to September 2016, leaf trait sampling took place at three of our studied elevational sites (0, 500, and 1500 m a.s.l.). In this field campaign, we aimed to resample as many vascular epiphyte species from the first survey as possible. At each elevation, epiphytes were sampled up to a height of 20 m on one or more trees using the single-rope climbing technique. Epiphytes below 6 m were sampled from the ground using a collecting pole. Functional traits were collected for all vascular epiphyte species classified as holoepiphytes (epiphytes in the strict sense, i.e. living their whole life cycle as epiphytes). In this dataset, we excluded nomadic vines because of their contact to the ground (Zotz 2013). Additionally, we excluded species of the family Cactaceae from trait measurements because stems are their main photosynthetic organs. This dataset differs in the sampling resolution between morphological and chemical traits; morphological traits include leaf measurements per individual at each study site and chemical traits include one measurement (from pooled samples) per species from each study site.

Leaf trait measurements
We collected between one and three leaves per adult individual from three individuals to obtain, if possible, a maximum of 10 leaves per species. We sampled fully expanded leaves without visible signs of herbivory or disease. Collected leaves were rehydrated in a sealed plastic bag and kept cool in a refrigerator at 7 °C for a minimum of 8 hours before taking measurements. , ii) specific leaf area (SLA = leaf area/dry weight; mm mg ), iii) leaf density (LD = SLA/leaf thickness; g cm ), and iv) leaf dry matter content (LDMC = dry weight/fresh weight; g g ) (Suppl. material 3, Fig. 5).
We measured the following leaf chemical traits: i) leaf nitrogen content (leaf nitrogen; %), ii) leaf carbon content (leaf carbon; %), iii) leaf phosphorus content (leaf phosphorus; %), iv) nitrogen isotope ratio (d N; ‰), and v) carbon isotope ratio (d C; ‰) (Suppl. material 4, Fig. 6). Dried leaf samples were ground and homogenized using a ball mill. To quantify leaf nitrogen content, leaf carbon content, d N, and d C, we used an elemental analyserisotope ratio mass spectrometer (Carlo Erba 1110 EA coupled via a Conflo III to a Delta ; Thermo Electron, Bremen, Germany). We used an internal standard, which is a solution of proline and sucrose with a C:N ratio of 8.8, d N of 0.16 (+/i 0.15) and d C of -10.20 (+/-0.13). We tested standards every ten samples, after which the IRMS was recalibrated using five certified isotope standards, i.e. IAEA-600, IAEA-N-1, IAEA-N2, and USGS-25. Atmospheric air (AIR) was used for d N and the Vienna Pee Dee Belemnite (V-PDB) for d C as standards.  1985). Leaf phosphorus concentrations were determined colorimetrically (Murphy and Riley 1962). After digestion, 770 μl distilled water was added and the absorption by the molybdenum-phosphorous complex was measured at 710 nm using a UV-VIS spectrophotometer (Specord 50, Analytik Jena, Jena, Germany). Chemical analyses of samples were performed at the University of Oldenburg for phosphorus and at the University of Vienna, Department of Microbiology and Ecosystem Science for nitrogen, d N, and d C.

Species identification
Vouchers from the first field campaign were collected, if possible, in triplicate for preservation as herbarium specimens. These specimens were identified using relevant literature ( (14), and Araceae (12). 72.2% of the sampled epiphyte individuals could be identified to species level, while another 26.1% were identified to genus level, and 1.7% to family level. The trait data set includes measurements for 1595 leaves from 474 individuals belonging to 102 species in 10 families. In total, most species were orchids (42.7%), followed by ferns (28.1%), and bromeliads (20.4%).
2) Phorophytes: The 120 climbed host trees belong to 32 tree species distributed in 25 genera and 21 families. Tree identification to the species level was possible in 53% of the cases, while another 44% were identified to genus level and 3% to family level.

Data set name: Species names
Description: Species scientific name and its corresponding family and species code (Suppl. material 5).

Column label Column description
Species.code Code for each scientific species name Species.name Scientific name of the species

Family
Family of the species

Additional information
We provide the description of the content and structure of each supplementary material in Table 1, with the source of standardization for each term used according to Darwin Core glossary and the Thesaurus of Plant characteristics.
Carbon isotope ratio (d C;‰) Carbon isotope ratio (d C;‰) The ratio of C to C of a leaf. ‰