Plant distribution and modern pollen deposition across an elevation eco-gradient: The lesson learnt from a case study in the Italian Alps

Abstract

This study investigates the distribution of vegetation and its modern pollen representation along an elevation gradient in the Italian Alps and explores the relationships with terrain and climate variables. Moss polsters were collected at 25 sites between ca. 300 and 1400 m asl from open areas, deciduous, and conifer forests. At each site vegetation was surveyed at 1.8 and 10 m radius according to the Braun-Blanquet method. Climatic data, bioclimatic indices, and terrain parameters were obtained for each sampling site. Three distinct pollen associations reflect the characteristic vegetation altitudinal belts present in the study area. Uphill dispersal, the regional load and the presence of high producers influence the abundance of pollen and the representation of taxa along the gradient. CCA ordination technique reveals the predictive power of environmental variables on modern pollen and vegetation datasets. The most relevant factors controlling vegetation distribution are identified in elevation, insolation, Pspring, Tsummer, the Ellenberg quotient, and Summer Water Balance (SWB). Elevation, insolation, summer P and T, the Ellenberg quotient, and the Gams’ hygric continentality index (GAMS) explain more variance within the pollen dataset. A qualitative comparison among pollen and the corresponding parent plant occurrence qualifies Abies alba, Fagus sylvatica, Poaceae and Cyclamen pollen as suitable indicators taxa of the local vegetation in the study area. Several high producers with very effective, long-distance anemophilous dispersal (Pinus, Ostrya, Alnus, and Juglans) or mixed pollination mechanisms (anemophilous and insect-pollination: Fraxinus ornus, and Castanea) show no or little association with their parent plants.

Keywords

Elevation gradient environmental parameters modern pollen deposition Southern Alps vegetation altitudinal belts

Introduction

The distribution of plants according to elevation always fascinated explorers and natural scientists, but we owe its first impressive representation to Alexander von Humboldt. During his epic ascent to the top of the Chimborazo Volcano (Ecuador) in 1802, he conceptualized “nature as a living system, where all processes depend on interactions and reciprocity.” After that trip, he painted the well-known Naturgemälde (von Humboldt and Bonpland, 1807), considered the first infographic ever realized: a topographic cross-section of the volcano showing relationships between elevation and plant distribution. von Humboldt sensed that altitude was only one of the controlling factors, and provided indications on atmospheric pressure, temperature, light, and geology at the most striking boundaries.

From the macroscopic (vegetation) to the microscopic (pollen) level, it is becoming more and more relevant for paleoecological and biodiversity conservation studies to investigate the relationships between the plants growing in a certain region and pollen deposition documented by natural or artificial traps. In the past decades, eco-gradients were traced in different biomes in Africa (among others, Bonnefille et al., 1993; Julier et al., 2018; Schüler et al., 2014; Tabares et al., 2018), America (Castro-López et al., 2021; Correa-Metrio et al., 2011; De Oliveira Portes et al., 2020; Urrego et al., 2011), Asia (Guo et al., 2020; Huang et al., 2018; Quamar et al., 2021; Zhang et al., 2018), and Europe (Connor et al., 2021; Fall, 2012; Finsinger et al., 2007; Furlanetto et al., 2019; Hjelle, 1999; Morales-Molino et al., 2020; Ortu et al., 2010; Senn et al., 2022; Servera-Vives et al., 2022). Most of these papers highlight the relationship of both, vegetation and pollen deposition with elevation, others with land use. Assuming that elevation largely influences the development of plant ecosystems (according to a mean elevation temperature lapse rate of −0.65°C/100 m altitude for standard atmosphere: International Civil Avian Organization, 1993) and likely their pollen image, we must be aware that other factors, often collinear, might play a relevant role. Recent studies (Zhang and Steiner, 2022 and references therein) suggest that pollen production is directly influenced by climate, as shown by the positive relationships of warmer temperatures and shifts in the start of spring dispersal and its longer duration.

This paper aims to fill a gap of knowledge in montane regions concerning the influence that climate and terrain parameters exert on vegetation and pollen deposition. We developed an elevation eco-gradient in the two sides of a montane area of the Italian Alps to test the effects of terrain and climate variability typical of opposite slopes on plant distribution, pollen deposition, and local climate and microclimate, respectively.

This paper aims to: (i) to assess gradients of vegetation distribution and modern pollen deposition in the study area, (ii) to understand the relationships between macro- and microscopic biodiversities and terrain/climate parameters, and (iii) to define suitable and non-suitable plant indicator taxa according to the degree of association between pollen and their parent plants.

The study area: An overview on geology, soils, climate, and vegetation

The elevation gradient presented in this paper develops in the south-eastern part of the Trentino Province (Northern Italy), crossing the Folgaria and Lavarone Highland (Figure 1a). The study area is bordered to the north by the Valsugana Valley, to the west by the Adige Valley, to the east and south by the Province of Vicenza. The Highland has a mean elevation of 1000–1200 m asl and it is surrounded by peaks reaching 2000 m asl. The area is part of a system of low-elevation limestone reliefs, known as “Vicenza Prealps.” Three main geological formations are mapped, that is Calcari Grigi, Maiolica and Dolomia Principale. Moraine deposits of phyllites, porphyries, and gneiss occur (Bertola and Cusinato, 2004), as well as glacial and karstic depressions, hosting the Lavarone Lake and several bogs (Pedrotti and Venanzoni, 1987). Soils developed on alluvial deposits and fans are classified as Rendzic Leptosols, Haplic Phaeozems, and Haplic Regosols; limestone and dolomite-rich bedrock hosts Rendzic Leptosols, Rendzic to Haplic Phaeozems and Cutanic Luvisols (Sartori and Mancabelli, 2009).

Figure 1.

(a) Map of the elevation transect superimposed on a simplified map of regionally-developed vegetation belts. Sampling sites are indicated with green, yellow and red dots. (b) 3D Sketch showing the morphology of the study area and the extent of vegetation belts.

Climate in the highland is sub-oceanic to subcontinental while the eastern and western valleys record a sub-Mediterranean influence, typical of the warm lowlands (Odasso et al., 2018). The mean annual precipitation (1990–2019) is 1342 mm, while mean temperatures (1990–2020) range on average from 18.0°C in summer to 1.3°C in winter (mean values calculated from the climatic series used in this study).

Based on the elevational arrangement of vegetation, three main belts were recognized in the study area (Figures 1 and 2 and Supplemental Table 1, available online). A colline belt extends from the valley floors up to 600 m asl, characterized by a sub-Mediterranean Ostrya carpinifolia forest with Fraxinus ornus, Fagus sylvatica, Tilia cordata, and Acer campestre. Pinus nigra and Pinus sylvestris forests are found on the driest sites. Vineyards, orchards, and cultivated fields are common and locally abundant. A sub-montane belt covers the range 600–900 m asl. Woodlands are dominated by a thermophilous beech forest of the Eastern Alps admixed with Fraxinus ornus, Ostrya carpinifolia, Corylus avellana, Populus tremula, and conifers such as Pinus sylvestris and Larix decidua. The Fagus sylvatica forest hosts sporadic trees of Abies alba and Picea abies from 700 m asl on shady slopes. Scrubs of Pinus mugo and Rhododendron hirsutum (Rhododendro hirsuti-Pinetum mugo: Pignatti and Pignatti, 2013) are locally present on steep rocky terrain and fens. A narrow layer of beech forest mixed with spruce and fir marks the transition to the montane belt at ca. 900 m asl. This belt covers the entire Lavarone Highland and reaches up to 1300 m asl. Human disturbance is here visible by the presence of ski infrastructures, hotels and private households, and pastures often abandoned and replaced by secondary forest. The vegetation is dominated by mixed forests of spruce, beech, and fir with an understory of graminoids and dwarf shrubs of Vaccinium myrtillus (Eastern Alpine Abies alba-Fagus sylvatica forest: Pignatti and Pignatti, 2013). Where anthropogenic pressure is higher, the landscapes are characterized by mown meadows (Arrhenatherion elatioris, Polygono-Trisetion flavescentis) and grazed grasslands (Cynosurion cristati) surrounded by mixed forests of beech and artificial stands of spruce. Patches of wet meadows (Molinion caeruleae) and transition mires with Carex rostrata are relatively common in the highland.

Figure 2.

Extent of the main forest types in the study area. (a) Abies alba (fir) – dominated forest types developed on calciphilous and mesic soils. (b) xeric Pinus sylvestris/P. nigra (pines) forest types on calciphilous soils. (c) Picea abies (spruce) forest types: montane forests and secondary formations, where the abundance of spruce is the result of forest management that favored its spread without any intentional introduction of the species. (d) Fagus sylvatica (beech) forest types, including beech-dominated submontane forests, mixed beech-conifer forests and pure beech stands. (e) Fraxinus ornus/Ostrya carpinifolia (manna ash/hop hornbeam) forest types, admixed with deciduous species of Quercus (oaks), Sorbus sp. (rowan) and few conifers. (f) Open/young forest types: shrubs-dominated grooves with Betula sp. (birches), Prunus sp. (plums), Sorbus sp. (rowan), Salix sp. (willows) and invasive species such as Ailanthus sp. (tree of heaven), Paulownia tomentosa (princess tree), and Acer negundo (boxelder).

Materials and methods

Sampling design

The elevation transect stretches over a considerable horizontal distance (almost 30 km) and consists of 25 sites sampled in June 2021. Sites are distributed along a NE – W route from 524 m asl in Lochere to 287 m asl in Calliano (Figure 1a). Each sampling site is indicated by an acronym composed by two letters and a number, listed in Supplemental Table 1, available online. The highest sampling site (acronym LV10) is located on the Lavarone Highland at 1329 m asl (Passo Sommo). At each sampling site, two to three moss polsters were collected within an area of 1.8 m radius (ca. 10 m²) and mixed into one sample. The geographical coordinates of each sampling sites were GPS recorded, terrain parameters determined using a DTM in ArcGIS 10.8 (ESRI) and climate variables obtained from climatic series of the last 30 years (1990–2019).

Vegetation was described at each sampling sites following the Braun Blanquet method, using a semi-quantitative abundance-dominance index to estimate the plant cover for each taxon in a radius of 1.8 m (ca. 10 m²) and 10 m (ca. 314 m²) around the sampling site. The value “R” was assigned to taxa with low occurrence (rare, 1–2 individuals), “+” (2–5 individuals), “1” for occurrence less than 5% coverage, “2” between 5% and 25%, “3” (25–50%), “4” (50–75%), and “5” (75–100%).

At a broader scale, the main forest types occurring in the study area were mapped in ArcGIS 10.8 (ESRI) using data from the Forest Service of the Trento Province (Provincia Autonoma di Trento – Servizio Foreste e Fauna, Regione del Veneto – Unità Organizzativa Foreste e Selvicoltura) (Figure 2).

Microbotanical analysis

Moss polsters were processed to concentrate microbotanical remains at the Laboratory of Palynology and Palaeoecology of CNR-IGAG in Milano. After weighting and volume estimation, Lycopodium tablets were added to each sample to allow the calculation of pollen and pollen-slide charcoal particles. Mosses were soaked in boiling KOH 10% to remove cellulose and humics, and thoroughly washed through a 250 µm sieve. The liquid was then put in vials and centrifuged to concentrate the microscopic content. Samples were then treated with 39% HF and 10% HCl to remove excess of mineral particles. Samples were finally acetolyzed, sieved at 7 µm and then glycerine was added to allow sample preservation and slide preparation. Slides were analyzed under a NEXCOPE NE610 light microscope at x400 and x630 magnifications. Pollen grains were identified using photographic atlases and dichotomous keys such as Beug (2004), Moore et al. (1991), Punt et al. (1976–2009), Reille (1992–1998), and the CNR-IGAG modern reference collection. The identification was made at the lowest possible taxonomic level. The identification of certain taxa followed the criteria presented in Furlanetto et al. (2018). Pinus cembra is differentiated from Pinus sylvestris/mugo type by the presence of verrucae on the ventral surface of the grain body, by the elongated shape of the sacci and by the occurrence of an ondulating outline in the proximal surface of the grain body. Based on grain size and pore + annulus diameter, three groups of Poaceae are distinguished. Grains with a diameter <37 µm are referred to wild Poaceae, grains with diameters between 37 and 47 include mostly wild grasses and some cultivated ones (Panicum sp.), while grains >47 µm and pore + annulus diameter >11 µm have been referred to cereals (Avena, Triticum group and Secale). Spores, pollen-slide charcoal particles and other microscopic biological particles were recorded, too. Charcoal particles were differentiated into two size classes, 7–50 µm and 50–250 µm length. A minimum of 600 pollen grains were counted per sample. Aquatics, spores, and non-pollen palynomorphs were excluded from the pollen sum used for percentage calculations, which includes trees, shrubs, upland herbs, and anthropogenic/cultivated plants. Pollen diagrams were drawn using TILIA 2.0.41. A selection of taxa is reported in Figure 3.

Figure 3.

Synthetic pollen – vegetation diagram. (a) The upper diagram displays conifers, deciduous taxa and shrubs. (b) The lower diagram depicts upland herbs, xerophytes, and pollen and charcoal concentrations. The vertical scale is uneven for a better visualization of closer samples between 1100 and 1300 m asl.

Terrain and climate parameters

Terrain parameters were extracted from a high resolution Digital Terrain Model (DTM) with x, y, z units (grid cell size 1 m × 1 m) provided by the Trentino-Alto Adige regional service (LIDAR PAT, 2014, updated to 2018). The elaboration was performed using ArcGIS 10.8 version (ESRI) to obtain mean values of elevation, aspect, slope, and insolation for each sampling site (ArcGIS, 2020). The calculated parameters are listed in Supplemental Table 2, available online. Aspect identifies the surface direction at each location, measured clockwise in degrees from 0° to 360° (both due North), while −1 indicates flat areas. Slope represents the steepness degree of the surface, with low values assigned to flat terrains and high values to sharp terrains. Insolation measures the solar radiation over a certain location, expressed in units of watt hours per square meters (Wh/m²).

Site specific monthly temperatures and precipitation series, covering the period 1990–2019, were reconstructed for each sampling site by means of the anomaly method (Mitchell and Jones, 2005) and exploiting the huge amount of instrumental data available for the area. The spatial distribution of the stations used for the reconstruction of climate parameters for each sampling site is shown in Supplemental Figure S1, available online. Specifically, as described in Brunetti et al. (2012, 2014) and Crespi et al. (2018) the technique is based on the independent reconstruction of monthly climatologies (i.e. the mean climate values estimated over a specific reference period) and on their deviations (i.e. anomalies with respect to the same baseline period). Climatologies are characterized by strong spatial gradients that need many weather stations (even if available for a short period) to be properly captured, together with an interpolation technique exploiting the dependence of climate variables on geographical parameters. Anomalies, linked to climate change and variability, present higher spatial coherence and a limited number of stations is enough to capture spatial patterns through simpler interpolation methods, but data homogenization is mandatory: the data must be corrected for errors deriving from changes in station and instrument location, instrument replacements, etc., which mask the actual climate signal. Finally, monthly temporal series can be obtained by superimposition of the reconstructed climatologies and anomalies. Mean monthly potential evapotranspiration (PET Penman Potential evapotranspiration, expressed as kg × m²/month) for the period 1990–2019 was extracted for each sampling site from the CHELSA high resolution database (Karger et al., 2017, 2020). Four bioclimatic indices (Ellenberg quotient, GAMS, CWB, SWB) were also calculated following Gams (1932), Noce et al. (2020), and Kasper et al. (2022). The Ellenberg quotient relates the temperature of the warmest month of the year to the annual precipitation and it is regarded as a simple measure of humidity in respect to continentality; it is expressed in °C/mm as the ratio Tjuly/Pann multiplied by 1000. Gams’ hygric continentality index (GAMS) was calculated as arccotg (Pann/elevation). The annual climatic water balance (CWB) and the mean summer water balance (SWB, i.e. the mean of June, July, and August) were calculated by subtracting mean annual or summer monthly potential evapotranspiration (PET) from mean annual or summer monthly precipitation.

Terrain and climatic data used for this study are presented in Supplemental Table 2, available online.

Multivariate analysis

Prior to the statistical analysis, modern pollen and vegetation data over 2% of abundance were selected and the percentage values were square-root transformed for variance stabilization. All computations were made using R 4.1.1. version (R Core Team, 2021) and the vegan package (version 2.5-7; Oksanen et al., 2016).

Detrended correspondence analysis (DCA) was applied to describe the major gradients in the modern pollen and vegetation assemblages and to examine similarities between vegetation composition and its pollen representation (Figure 4). Canonical correspondence analysis (CCA) was used to evaluate the effects of terrain and climatic parameters on pollen and vegetation abundances. The explanatory power of each variable (14 predictors were considered, see Table 1) was first assessed individually and then, a summary CCA analysis was performed using six main predictors concurrently (Figure 5).

Table 1.

Results of canonical correspondence analysis (CCA).

Pollen
	Terrain parameters				Climate data										Bioclimatic indices
	elevation	insolation	aspect	slope	Pann	Pspring	Psummer	Pautumn	Pwinter	Tann	Tsummer	Twinter	Tjuly	Tjan	Ellenberg	GAMS	CWB	SWB
variance explained	13.05%	5.63%	4.70%	2.60%	11.42%	11.83%	13.56%	11.49%	7.28%	12.46%	12.61%	11.24%	12.61%	12.16%	12.41%	11.35%	7.80%	6.36%
p	0.001***	0.181	0.308	0.838	0.004**	0.003**	0.001***	0.001***	0.057	0.002**	0.002**	0.003**	0.002**	0.002**	0.003**	0.001***	0.049*	0.099
Vegetation 10 m radius
	Terrain parameters				climate data										bioclimatic indices
	elevation	insolation	aspect	slope	Pann	Pspring	Psummer	Pautumn	Pwinter	Tann	Tsummer	Twinter	Tjuly	Tjan	Ellenberg	GAMS	CWB	SWB
variance explained	6.16%	7.01%	4.44%	3.75%	6.93%	6.96%	6.88%	6.88%	6.84%	6.06%	6.10%	5.75%	6.10%	5.96%	6.33%	5.73%	7.40%	7.51%
p	0.003**	0.001***	0.297	0.682	0.001***	0.001***	0.001***	0.002**	0.001***	0.009**	0.003**	0.022*	0.008**	0.007**	0.002**	0.013*	0.001***	0.001***
Vegetation 1.8 m radius
	Terrain parameters				Climate data										bioclimatic indices
	elevation	insolation	aspect	slope	Pann	Pspring	Psummer	Pautumn	Pwinter	Tann	Tsummer	Twinter	Tjuly	Tjan	Ellenberg	GAMS	CWB	SWB
variance explained	6.40%	6.44%	3.89%	4.21%	6.43%	6.57%	6.49%	6.32%	6.23%	6.43%	6.42%	6.41%	6.42%	6.45%	6.47%	6.38%	6.91%	7.03%
p	0.003**	0.001***	0.634	0.483	0.001***	0.001***	0.001***	0.001***	0.007**	0.002**	0.003**	0.006**	0.003**	0.006**	0.002**	0.003**	0.001***	0.001***

Significance codes: ***0.001; **0.01.

Proportions of variance explained by each variable and permutational-repeated measures analysis (ANOVA) obtained for pollen and vegetation data (surveys at 1.8 and 10 m radius), abundance ≥2% in at least one sample. Bold term represents the six main predictors used concurrently for the summary CCA analysis (see Figure 5).

Figure 4.

Detrended Correspondence Analysis (DCA) biplots of pollen and vegetation percentages. (a) modern pollen deposition data, (b) vegetation data within 10 m radius, (c) vegetation data within 1.8 m radius. Dissimilarities between pollen and vegetation assemblages can be appreciated with DCA.

Figure 5.

Canonical correspondence analysis (CCA) biplot of pollen and vegetation percentages and terrain/climatic parameters (a) modern pollen deposition data, (b) vegetation data within 10 m radius, (c) vegetation data within 1.8 m radius. The first five variables showed significant and independent contributions to the calibration set variance, whereas Tsummer is collinear with elevation. The variable explaining less variance is shown with a dotted line.

Results

Modern pollen assemblages and comparison with vegetation data

The most significant and frequent taxa recorded in modern pollen spectra and vegetation in 10 m radius surveys are plotted in Figure 3 along an y axis representing elevation (m asl). About 146 types of microbotanical proxies (pollen, spores, and other biological particles) were identified in the analyzed samples. 204 plant and 4 fern species were detected in vegetation surveys within a 10 m radius.

Pollen spectra are dominated by big pollen producers such as Pinus sylvestris/mugo type, Fraxinus ornus, Ostrya, and Corylus. Picea pollen is lacking in several samples below 1000 m asl, while occurring continuously in mosses sampled next to spruce stands (Figure 2). Despite the abundance of Abies alba in the study area (Figure 2), plants were found only at two sampling sites (LV1 and LV3). Fir is a moderate pollen producer, but its dispersal is poor due to the very large size of pollen grains, resulting in a high terminal velocity (fall speed) and deposition in a limited radius around the parent plants. On the contrary, plants of Ostrya carpinifolia and Corylus avellana were rarely encountered along the trail but their pollen is abundant in almost all analyzed samples. Both plants are huge pollen producers with very effective dispersal mechanisms. Fagus sylvatica is quite common and abundant both, at low altitudes and on the Lavarone Highland (Figure 2), but its pollen percentages in modern samples remain usually low (<5%), exceeding 20% only in one site on the Highland. Pollen of Fraxinus ornus is extremely abundant in modern samples from 600 to 1000 m asl. This pollen feature is not mirrored by the occurrence of ashes in modern vegetation.

Pollen of deciduous Quercus species, Carpinus betulus, Alnus glutinosa type (including Alnus glutinosa and A. incana), Castanea, and Juglans appear almost in all samples, despite their absence on very local vegetation and limited occurrence at larger scale. Among herbs, Poaceae, Ranunculus acris, Plantago lanceolata, and Artemisia are quite common in pollen spectra, but only grasses are regularly recorded in vegetation.

Beside pollen, other microscopic particles (not shown in Figure 3) are frequent in the analyzed samples, that is, the Thecamoebae Arcella and Assulina, fungal spores of the families Sordariaceae and Xylariaceae (Ustulina) and fern spores of the genera Asplenium and Pteridium.

Small charcoal particles (7–50 μm length) are recorded in most of the analyzed samples, although in very low concentrations, with sample LV1 as only exception. This site is next to some holiday apartments and charcoal particles are likely the result of domestic activities. Larger charcoal particles (50–250 μm length) occur in two samples only (VS5 and LV8) in very limited amounts.

Detrended correspondence analysis (DCA)

Our elevation gradient shows differences between pollen abundance and plant representation, likely caused by the homogenization effect of wind dispersal combined with the occurrence of high pollen producer taxa (Figure 4). The DCA on modern pollen assemblages (Figure 4a) is characterized by shorter axis length (DCA1 axis length 1.3) than vegetation assemblages (Figure 4b and c; DCA1 axis length 4.5 in 10 m radius and 3.6 in 1.8 m radius). This difference depends on the different taxonomic accuracy used for pollen and vegetation data. Several plant families cannot be resolved at lower levels by palynological analysis, while genus and species level can easily be reached by plant identification. In the modern pollen assemblages, sites from the three different sections of the transect (i.e. Adige Valley – VA, red dots, Valsugana Valley – VS, green dots and the Lavarone Highland – LV, yellow dots) are well subdivided along axis 1 (Figure 4a). Axis 1 likely represents the elevation gradient and the greater pollen richness documented in herb taxa, actually corresponding to a larger number of herbs identified in open forests on the Lavarone Highland. In vegetation assemblages, sites from the three sections of the transect are instead more combined (Figure 4b and c).

Canonical correspondence analysis (CCA)

A total of 18 environmental variables were used for Canonical Correspondence Analysis (CCA). The variables and the variances explained by each predictor are listed in Table 1.

As for pollen assemblages, the most significant explanatory variable is summer precipitation (Psummer), accounting for 13.6% of the variance. Summer water balance (SWB) is the most significant predictor both for the 10 and 1.8 m vegetation composition, accounting for the 7.5% and 7.0%, respectively. The CCA ordination biplots of pollen and vegetation assemblages performed with six main predictors concurrently is presented in Figure 5. In Figure 5a, the total variance explained is 40.8 %. Axis 1 (CCA1) contrasts high elevation sites, from the Lavarone Highland (LV, yellow dots), with positive scores, with the Adige Valley (VA, red dots) and Valsugana Valley (VS, green dots) sites featuring negative scores. Sites VA1-5 are characterized by high values of Tsummer and Ellenberg quotient and high pollen percentages of Fraxinus ornus. In opposition sites LV1 and 10 are distinguished by high values of GAMS index, Psummer and elevation and high pollen percentages of Abies alba, Larix, and Picea. Axis 2 (CCA2) separates sites with high insolation (e.g. LV9, VA1), distinguished by high percentages of Artemisia, from sites with low insolation and high pollen percentages of Pinus sylvestris/mugo, Pinus cembra, and Alnus glutinosa type.

In the CCA plot of vegetation data within 10 m radius (Figure 5b), the six predictors explain 32.3% of the variance. Axis 2 (CCA2) separates high elevation sites (LV sites, negative scores) associated with Abies alba, Myosotis arvensis, and Trollius europaeus from low elevation sites (positive scores), marked by high values of Tsummer and Ellenberg quotient and high percentages of Ostrya carpinifolia and Pinus sylvestris. Axis 1, on the other hand, divides sites with high insolation and summer water balance (SWB) (Lavarone Highland and Adige Valley) from Valsugana Valley sites. VS6-7 sites reflect the transition from the submontane to montane altitudinal belt since they are located close to Fagus sylvatica and Picea abies.

In the CCA plot of vegetation data within 1.8 m radius (Figure 5c) the six predictors explain 32.1% of the variance in the vegetation. Axis 1 separates sites with high values of insolation and Pspring (Lavarone Highland and Adige Valley) from Valsugana Valley sites. Site LV9, located in an abandoned area in a rural context, is clearly distinct from the other sites, yielding high insolation values and specimens of Artemisia sp. and Crassulaceae. Sites LV11, LV10, LV7, LV5-4 show similar vegetation and are in partially treeless areas, bordering large grasslands. Axis 2 (CCA2) separates sites with high values of Tsummer and Ellenberg quotient (e.g. VA5-7 with negative scores) and high percentages of Epimedium alpinum, Lamium sp., and Cornus sp.

Discussion

Comparing modern pollen deposition with local to regional vegetation distribution

Distinctive modern pollen associations from moss polsters mirror the arrangement of vegetation in the study area. The vegetation altitudinal belt growing up to 600 m asl, composed by oak, pine, ash, and hop hornbeam (Figure 1b), finds a counterpart in the pollen image provided by sites VA7 to VA4 (Figure 3). The corresponding spectra are dominated by pollen of Pinus sylvestris/mugo type, Fraxinus ornus, and Ostrya, with minor contributions from other broad-leaved trees. The low representation of herbs pollen reflects the consistency of trees and shrubs in forest communities, limiting the development of a thick herb understory.

Mixed forests with beech, ash, hop hornbeam, and pine occur between 600 and 900 m asl (Figure 1b). The dominant tree in the 10 m radius surveys is Fagus sylvatica (Figure 3a), which currently occupies more than 19% of the total land surface in the study area (Figure 2d). Pollen spectra from sites VS2 to VA2 still record high percentage values of ash and hop hornbeam, accompanied by Fagus and decreasing, but still abundant, pine.

The highest pollen percentage values of Picea, Abies, and Fagus are recorded at sites above 900 m (VS5 to LV10, Figure 3a) mirroring the composition of mixed forests dominated by spruce, fir, and beech. Above 900 m the number and abundance of herb species in vegetation and their pollen percentages in moss polsters increase, reflecting increased forest openness and the development of a rich herb understory.

Acer and Juglans were rarely or never reported in the vegetation surveys, but their pollen is documented in some modern pollen spectra. This confirms that the relevant pollen source area – RPSA (Sugita, 1994) of certain taxa is, as expected, larger than the total single-plot area surveyed in this study (ca. 314 m²) and varies according to the vegetation structure and to the landscape topography.

We also detected the presence of pollen of Vitis and Olea (not shown in Figure 3), two allochthonous taxa that were not found in the vegetation surveys. Vineyards are quite common on the valley floors and small patches for domestic use were also observed in the rural areas of the highland. No evidence of olive trees was found in the study area. Potential sources of Olea pollen might be the drier valleys in the western part of the Trentino region close to Lake Garda, 30 km away from the study area (e.g. Rodrigo-Comino et al., 2021).

Our elevation gradient sometimes shows a mismatch between pollen abundance and plant representation, likely caused by the homogenization effect of wind dispersal combined with the occurrence of high pollen producer taxa (Broström et al., 2008; Court-Picon et al., 2005). Pollen mixing and pollen transport from external sources, due to regional or local wind circulation pattern, are frequent in montane areas (Markgraf, 1980). Single species might strongly dominate the pollen spectrum and obscure the representation of low pollen producers (Räsänen et al., 2007). These anomalies can be observed also in open areas where pollen production by herbs is usually low and the extra-local component increases (de Nascimento et al., 2015; Papadopoulou et al., 2022). In our pollen spectra, this effect is visible for Pinus sylvestris/mugo type and for deciduous species such as Corylus, Fraxinus ornus, Ostrya carpinifolia, and Fagus sylvatica. As for Corylus, its pollen productivity seems to be associated to the degree of forest openness and it tends to flower less within woods than in open vegetation (Broström et al., 2008). Moreover, its pollen dispersal is more difficult in closed vegetation, with pollen grains impacting on leaves and then useless for reproduction purposes.

The quantification of the percentage of land surface covered by forest types (Figure 2, total land surface corresponding to 249.25 km²) provides further indications on the potential of moss polsters to document local to regional vegetation, a topic with great impact for the interpretation of fossil pollen records.

Less than 17% of modern land surface, corresponding to 41.6 km², is covered by forest communities dominated by Abies alba (Figure 2a) and managed as high forest to promote a very heterogeneous structure through selective cutting. Fir is a moderate pollen producer, but its pollen dispersal is not effective (Poska and Pidek, 2010) and its pollen is usually underrepresented in modern samples (Pidek et al., 2013). This is nicely shown at site LV1 where more than 20% of the surveyed area is covered by fir, but only 3–5% of pollen found in mosses belongs to this species.

6.4% of modern land surface (15.9 km²) is covered by xeric pine forests on calcareous soils with Pinus sylvestris and Pinus nigra (Figure 2b) managed through clear cuts on small patches. Pines are huge pollen producers (according to Geburek et al., 2012, ten times higher than the other conifers:) with very effective anemophilous dispersal, usually resulting in an overrepresentation of this conifer. Actually, pine was documented only in four sites along our transect, but its pollen is continuously present, often among the most abundant taxa (Figure 3).

Picea abies forests extend over 13.6% of the land surface (ca. 34 km², Figure 2c), either as spruce-dominated stands or secondary formations where spruce is not planted but instead favored along with silver fir by forest management through selective or shelterwood cutting on small groups to promote regeneration of single trees or small groups. Mixed forests of spruce and silver fir of the montane belt are sometimes mature and have a high structural diversity (uneven-aged stands). Pure spruce stands frequently replace the natural mixed spruce-silver fir and beech forests. Spruce forests occur on the Lavarone Highland whereas at lower elevations, only sporadic trees were observed on shady slopes. Its large pollen production (Hicks, 1994; Markgraf, 1980) is further enhanced by a very effective dispersal mechanism bringing its pollen at large distance from forests (Figure 3, sites below 900 m).

Fagus sylvatica is likely the most common tree in the study area. Pure stands and mixed forests cover more than 19% of the land surface (48 km², Figure 2d). Beech forests are managed as a high-forest system in monospecific or mixed formations with conifers as well as coppice with a rotation period of 25–30 years or admixed with hornbeam and manna ash. Fagus is a moderate pollen producer. Simulations of beech pollen dispersal show a drop in pollen load already at 300 m from the closest stand (Poska and Pidek, 2010). Despite the abundance of this plant at some of our sampling sites, its pollen is always underrepresented (Figure 3a).

Fraxinus ornus and Ostrya carpinifolia mix with some other deciduous species and few conifers in forests of the colline belt, covering 5.5% of land surface (13.8 km², Figure 2e). These forests are managed as coppice forests with a relatively short rotation period, no longer than 25 years. Both plants are good pollen producers with anemophilous dispersal, and they are among the most abundant taxa found in modern pollen spectra, although they are not so common in our eco-gradient.

Open/young woody stands with oaks, chestnut, hornbeam, and some shrubs often mark the transition from cultivated fields and pastures to more closed forests. This forest type extends over less than 1% of the land surface (2 km², Figure 2f). A comparison with pollen percentage of these taxa in modern samples seems to indicate that their representation is overall correct.

Predictive power of environmental variables on modern pollen deposition and vegetation distribution

The CCA statistical elaborations pinpoints the variables explaining most of the variance included in the modern pollen and vegetation assemblages. Elevation, insolation, Tsummer (mean of June-July-August monthly temperatures), Psummer for pollen, and Pspring (sum of March-April-May monthly precipitation) for vegetation are the most significant parameters (Figure 5 and Table 1). Among terrain parameters, aspect and slope exert slightly and statistically poor effects on our datasets. Their vectors remain close to the origin in all the preliminary CCA plots and were therefore excluded from the final elaborations.

Sites at high elevation on the Lavarone Highland (LV 1–11, yellow dots in Figure 5) almost always display their best fit on the positive side of axis 1 of the CCA plots where elevation and insolation vectors are longer, accompanied by Psummer and GAMS (GAMS hygric continentality index) for pollen deposition data (Figure 5a) and by Pspring and SWB (Summer Water Balance) for vegetation data (Figure 5b and c). On the contrary, sites at lower elevation (most of VS and VA sites, green and red dots in Figure 5) mostly have their best fit on the negative side of the CCA biplots where Tsummer and the Ellenberg quotient are the most striking parameters. The colline belt sites (VS1 in green and VA4 to 7 in red), located at the elevation gradient extremes, form quite homogeneous groups in Figure 5b and c, underlying their similarity in climate and vegetation.

Pollen assemblages seem to be strongly related to summer precipitation (Figure 5a). A possible explanation is that the pollen load suspended in the air is slumped to the ground by precipitations, occurring mostly during the summer months. Precipitations are also responsible for the mobilization of organic and mineral debris deposited on the ground (surface runoff processes).

The position of certain taxa in relation to the selected environmental variables (CCA biplots on the right side of Figure 5) allows deriving some ecological considerations (Gaillard et al., 1992). The signals extracted from the CCA biplots reflect the realized niche of certain plants in the study area and species distribution seems to be the expression of both biotic (competition) and abiotic factors (i.e. sum of spring precipitation, summer water balance, insolation, elevation). Ostrya carpinifolia and Pinus sylvestris are associated to high Tsummer and high value of the Ellenberg quotient, characteristic of south-facing slopes. Both species are light-demanding and preferably grow on warm, sunny or partially shaded slopes.

The ecological interpretation for Fagus sylvatica is more challenging due to its widespread distribution in the study area, from the dry colline belt to the more humid montane belt. Its intermediate position in all CCA biplots shows association with the GAMS continentality index (pollen), the Ellenberg quotient (vegetation 10 m radius) and Tsummer (vegetation 1.8 m radius). This finds a counterpart in the ecological requirements of this species and with its ecological limitations, as it preferentially grows on shaded areas and does not tolerate prolonged drought conditions and low temperatures (Leuschner, 2020).

In both pollen and vegetation (10 m radius) CCA biplots, silver fir (Abies alba) is associated to elevation. Association to Pspring is highlighted for the vegetation data. These features indicate that its occurrence is restricted to cooler and moister sites, where summer temperature is not too high.

Spruce (Picea abies) is a common species in the Lavarone Highland. In pollen and vegetation (10 m radius) CCA biplots this species is associated to elevation. Its sensitivity to water availability and spring temperature is suggested by the vegetation data (1.8 m radius) CCA biplots; spruce is generally a shade-tolerant species, but often occurs on sunny slopes with high water availability.

Indicator taxa

We compared the pollen percentages and the corresponding parent plant abundance for each site along the elevation eco-gradient (Figure 6), to discriminate when a taxon found in pollen is or is not a good indicator of local vegetation. These considerations are particularly relevant in view of the interpretation of fossil pollen records, where the local occurrence of a certain plant cannot be usually inferred solely on pollen data. Based on the visual degree of association between pollen and the occurrence of its parent plant in the local vegetation (Figure 6), we were able to distinguish “suitable” and “not suitable” indicator taxa. Our findings have been compared with several studies focusing on pollen productivity estimates (PPE) and on relevant source area for several plant taxa (Broström et al., 2008; Bunting et al., 2004; Calcote, 1995; Poska and Pidek, 2010; Räsänen et al., 2007; Sjögren et al., 2008b). It is important to note that pollen productivity may vary at the regional scale and year to year according to the age of the plant, the regional climate, and also the local microclimate (Sjögren et al., 2008b). It is also influenced by land-use practices, by vegetation structure, and growth forms (Broström et al., 2008).

Figure 6.

Comparison between plant abundance and pollen percentages along the elevation gradient, useful for the identification of suitable and not suitable indicator taxa.

Suitable indicator taxa

Two trees and two herbs can be qualified as suitable indicator of the neighboring vegetation, as pollen abundance and parent plants occurrences share a similar elevation pattern. Pollen of Abies alba is indicative of the local presence of this tree and marks the transition to the montane altitudinal belt in our study area. This result matches with the outcomes of other studies focusing on fir pollen productivity and dispersal (Poska and Pidek, 2010; Sjögren et al., 2008a) and confirms that pollen dispersal is relatively limited.

Although the distribution of Fagus sylvatica is not limited to one altitudinal belt, there is a moderate association between the pollen and parent plants in almost all plots. Our pollen spectra show that there is a minimal background load from beech pollen and that the highest pollen values are found in the elevation range where beech is most abundant.

Cyclamen qualifies as one of the best indicator taxa. There is a very good relationship among pollen and the parent plant occurrence in our elevation gradient, where Cyclamen plants were observed and relatively abundant at several sampling sites. Its peculiar flower structure is adapted to entomophilous pollination, mainly by flies and bumblebees (Affre and Thompson, 1997). Insect-pollinated taxa are usually under-represented in pollen spectra (Hjelle, 1998, 1999) due to their low pollen production. Our investigation shows that Cyclamen produced a considerable signal in our pollen assemblage.

The degree of association for Poaceae is quite high. Plants belonging to this family are high pollen producer and yield a diagnostic value for open herbaceous environments (Broström et al., 1998), sometimes biased by human practices (e.g. mowing; Hjelle, 1998). In our elevation gradient, the relationship between pollen of Poaceae and their occurrence in vegetation is particularly good, especially in sampling sites close to grasslands or forest edges.

Not suitable indicator taxa

The following pollen types are considered not suitable indicators for the local occurrence of plant, given their very low or near zero degree of association. Pollen of Ostrya carpinifolia and Corylus occur in all modern pollen assemblages while plants were rarely found within the surveyed vegetation. They both lack of diagnostic value due to their great dispersion capacity and high PPEs.

Pinus sylvestris/mugo type pollen is very abundant, and sometimes dominates, modern deposition assemblages. We observed a weak correlation with its parent plants (Pinus sylvestris, P. mugo, P. nigra). Pine pollen occurs also in open areas, above its primary vegetation belt. Its pollen grains easily disperse in the neighboring elevational belts and is well-known to contribute to the regional pollen rains in other biogeographic regions (Broström et al., 2008; Jackson and Wong, 1994; Papadopoulou et al., 2022; Poska and Pidek, 2010; Räsänen et al., 2007; Soepboer et al., 2007).

Pollen of Castanea, Juglans, and Alnus glutinosa type is frequently recorded in the modern pollen deposition although, parent plants were never found. They fall within the category of “Allochtonous pollen types” as defined by Cañellas-Boltà et al. (2009) in their altitudinal transect in the central Pyrenees. Alder pollen is likely the signal of riparian vegetation communities growing along rivers and in wetter thickets not included in our surveys. Chestnut grows below 600 m in the western part of our trail, but plants were not encountered. Chestnut is also cultivated in Valsugana (eastern part of the transect): here, the so-called “Castanea Road” (Strada del Castagno) as a renowned touristic attraction with thematic routes and excursions. So, the occurrence of chestnut pollen in our modern deposition record likely reflects a regional pollen load (cp. Gottardini et al., 2004). As for walnut, its production in the whole region dramatically decreased in the past decades, from 150 ha in 1979 to ca, 20 ha in recent years (Provincia Autonoma di Trento, 2011). It is traditionally cultivated in a limited area in western Trentino, quite far from our trail (almost 60 km from the nearest sampling site). Given the moderate pollen productivity of Juglans and its poor dispersal mechanism biased by large and heavy pollen grains, it is unlikely that the pollen documented in our record comes from this far. A possible source for this pollen might be single plants grown outside rural houses and farms in the study area for fruits and a very valuable flour.

Conclusion and outlook

The elevation gradient analyzed in the Southern Alps highlights the relationships between plant distribution and their pollen representation in moss polsters and also the role played by climatic and terrain parameters on macro- and microbotanical biodiversity. The analysis of pollen and vegetation elevation patterns suggests that the species composition of the vegetation belts occurring in the area is overall correctly represented in their pollen images. As expected, the presence of a wide array of high pollen producers and the anemophilous dispersal of pollen across the elevation gradient lead to a homogenization effect and the increase of the RPSA (Relevant Pollen Source Area) of several types (i.e. Pinus sylvestris/mugo type, Ostrya, Fraxinus ornus, and Corylus).

Elevation addresses a large part of the variance observed in pollen and vegetation data but, as correctly sensed by von Humboldt (and later on confirmed by other scholars and researchers), other parameters correlated to elevation play a relevant role. Moreover, in our gradient the predictive value of elevation is high for pollen data but decreases to half this amount in vegetation data. The same pattern is observed for all the other parameters, with the exception of the SWB (Summer Water Balance) index, which is slightly higher for both vegetation datasets than for pollen. These features need to be further evaluated in other elevation gradients.

Beside its importance in documenting the relationships between vegetation and modern pollen deposition in a montane area, this study will also have further applications. Actually, our sampling sites are located along the elevation gradient that hosts Lake Lavarone, drilled in the past years for palaeo-environmental analysis (Arpenti and Filippi, 2007; Filippi et al., 2007). A pollen record covering the past ca. 18,000 years is available from this site. Original data produced by the present research will be integrated in the next release of the EMPD (Eurasian Modern Pollen Database, Davis et al., 2020), and used to establish transfer functions for the reconstruction of past climate variables (Chevalier et al., 2020; Vallè et al., 2019).

Further developments to the research design applied in this project might be:

The establishment of artificial pollen traps in the same sampling sites, to determine annual pollen accumulation rates in different landscape scenarios and to estimate pollen production fluctuations through time;

The identification of mosses and the estimation of their growth rate, to assess the species-specific pollen accumulation period. The question “how many years of pollen deposition are retained in a moss cushion?” is highly debated. While some authors assume an average accumulation period of 5 years (Bradshaw, 1981), others reduce this period to around 2 years (Caseldine, 1981; Cundill, 1991). More recent estimations indicate a time period between 1 and 4 years (Lisitsyna and Hicks, 2014). These estimates may vary according to the moss growth form, that is, mosses adapted to xerophytic habitats and climates might have reduced growth rates and thus, the same moss thickness likely represent longer time periods.

The identification of plant macro- (i.e. seeds, conifer needles) and micro-remains (i.e. conifer stomata) from surface sediments to investigate the local representation of some taxa and compare it with the modern pollen assemblage.

Supplemental Material

sj-docx-1-hol-10.1177_09596836221138325 – Supplemental material for Plant distribution and modern pollen deposition across an elevation eco-gradient: The lesson learnt from a case study in the Italian Alps

Supplemental material, sj-docx-1-hol-10.1177_09596836221138325 for Plant distribution and modern pollen deposition across an elevation eco-gradient: The lesson learnt from a case study in the Italian Alps by Valentina Fontana, Giulia Furlanetto, Paolo Bertuletti, Michele Brunetti, Stefan Zerbe and Roberta Pini in The Holocene

Footnotes

Acknowledgements

The authors thank Dr. Cesare Ravazzi for suggestions and comments on an early draft of the manuscript. Miss B. Ceresa is thanked for support during fieldwork, Mr. Davide Margaritora for data mining within the CHELSA high-resolution database. Dr. Giulia Furlanetto acknowledges support from the Dept. of Environmental and Earth Sciences of University Milano-Bicocca. This paper is a contribution to the CNR Research Line DTA.AD001.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Dr. Giulia Furlanetto acknowledges support from the Dept. of Environmental and Earth Sciences of the University Milano-Bicocca (research grant type A1 junior).

ORCID iDs

Giulia Furlanetto

Roberta Pini

Supplemental material

Supplemental material for this article is available online.

References

Affre

Thompson

(1997) Variation in the population genetic structure of two cyclamen species on the island of Corsica. Heredity 78: 205–214.

ArcGIS (2020) ArcGIS [GIS software] Version 10.8. Redlands, CA: Environmental Systems Research Institute, Inc.

Arpenti

Filippi

(2007) Evoluzione della vegetazione nei pressi del Lago di lavarone (TN) negli ultimi 2200 anni. Studi Trentini di Scienze Naturali Acta Geologica 82: 317–324.

Bertola

Cusinato

(2004) Le risorse litiche dell’Altopiano di Folgaria e il loro utilizzo a Riparo Cogola. Studi Trentini di Scienze Naturali, Preistoria Alpina 40: 107–123.

Beug

(2004) Leitfaden der Pollenbestimmung Für Mitteleuropa und Angrenzende Gebiete. München: Verlag Dr. Friedrich Pfeil.

Bonnefille

Buchet

Friis

, et al. (1993) Modern pollen rain on an altitudinal range of forests and woodlands in South West Ethiopia. Opera Bot 121(71): 84.

Bradshaw

RHW

(1981) Modern pollen-representation factors for woods in south-East England. Journal of Ecology 69: 45–70.

Broström

Gaillard

Ihse

, et al. (1998) Pollen-landscape relationships in modern analogues of ancient cultural landscapes in southern Sweden ? A first step towards quantification of vegetation openness in the past. Vegetation History and Archaeobotany 7: 189–201.

Broström

Nielsen

Gaillard

, et al. (2008) Pollen productivity estimates of key European plant taxa for quantitative reconstruction of past vegetation: A review. Vegetation History and Archaeobotany 17: 461–478.

10.

Brunetti

Lentini

Maugeri

, et al. (2012) Projecting north eastern Italy temperature and precipitation secular records onto a high-resolution grid. Physics and Chemistry of the Earth 40–41: 9–22.

11.

Brunetti

Maugeri

Nanni

, et al. (2014) High-resolution temperature climatology for Italy: Interpolation method intercomparison. International Journal of Climatology 34(4): 1278–1296.

12.

Bunting

Gaillard

Sugita

, et al. (2004) Vegetation structure and pollen source area. The Holocene 14: 651–660.

13.

Calcote

(1995) Pollen source area and pollen productivity: Evidence from forest hollows. Journal of Ecology 83: 591–602.

14.

Cañellas-Boltà

Rull

Vigo

, et al. (2009) Modern pollen—vegetation relationships along an altitudinal transect in the central Pyrenees (southwestern Europe). The Holocene 19(8): 1185–1200.

15.

Caseldine

(1981) Surface pollen studies across Bankhead Moss, Fife, Scotland. Journal of Biogeography 8: 7–25.

16.

Castro-López

Domínguez-Vázquez

Islebe

, et al. (2021) Modern pollen-vegetation relationships across a landscape mosaic in central México. Review of Palaeobotany and Palynology 289: 104362.

17.

Chevalier

Davis

BAS

Heiri

, et al. (2020) Pollen-based climate reconstruction techniques for late Quaternary studies. Earth-Science Reviews 210: 103384.

18.

Connor

van Leeuwen

JFN

van der Knaap

, et al. (2021) Pollen and plant diversity relationships in a Mediterranean montane area. Vegetation History and Archaeobotany 30: 583–594.

19.

Correa-Metrio

Bush

Pérez

, et al. (2011) Pollen distribution along climatic and biogeographic gradients in northern Central America. The Holocene 21(4): 681–692.

20.

Court-Picon

Buttler

de Beaulieu

J-L

(2005) Modern pollen–vegetation relationships in the Champsaur valley (French alps) and their potential in the interpretation of fossil pollen records of past cultural landscapes. Review of Palaeobotany and Palynology 135: 13–39.

21.

Crespi

Brunetti

Lentini

, et al. (2018) 1961-1990 high-resolution monthly precipitation climatologies for Italy. International Journal of Climatology 38(2): 878–895.

22.

Cundill

(1991) Comparisons of moss polster and pollen trap data: A pilot study. Grana 30: 301–308.

23.

Davis

Chevalier

Sommer

, et al. (2020) The Eurasian modern Pollen Database (EMPD), version 2. Earth System Science Data 12(4): 2423–2445.

24.

de Nascimento

Nogué

Fernández-Lugo

, et al. (2015) Modern pollen rain in canary Island ecosystems and its implications for the interpretation of fossil records. Review of Palaeobotany and Palynology 214: 27–39.

25.

De Oliveira Portes

MCG

De Forest Safford

Montade

, et al. (2020) Pollen rain–vegetation relationship along an elevational gradient in the Serra dos Órgãos National Park, southeastern Brazil. Review of Palaeobotany and Palynology 283: 104314.

26.

Fall

(2012) Modern vegetation, pollen and climate relationships on the Mediterranean island of Cyprus. Review of Palaeobotany and Palynology 185: 79–92.

27.

Filippi

Heiri

Arpenti

, et al. (2007) Evoluzione paleoambientale dal tardoglaciale a oggi ricostruita attraverso lo studio dei sedimenti del lago di lavarone (Altopiano di Folgaria e Lavarone, Trentino). Studi Trentini di Scienze Naturali: Acta Geologica 82: 279–298.

28.

Finsinger

Heiri

Valsecchi

, et al. (2007) Modern pollen assemblages as climate indicators in southern Europe. Global Ecology and Biogeography 16(5): 567–582.

29.

Furlanetto

Ravazzi

Badino

, et al. (2019) Elevational transects of modern pollen samples: Site-specific temperatures as a tool for palaeoclimate reconstructions in the Alps. The Holocene 29(2): 271–286.

30.

Furlanetto

Ravazzi

Pini

, et al. (2018) Holocene vegetation history and quantitative climate reconstructions in a high-elevation oceanic district of the Italian Alps. Evidence for a middle to late Holocene precipitation increase. Quaternary Science Reviews 200: 212–236.

31.

Gaillard

Birks

Emanuelsson

, et al. (1992) Modern pollen/land-use relationships as an aid in the reconstruction of past land-uses and cultural landscapes: An example from south Sweden. Vegetation History and Archaeobotany 1: 3–17.

32.

Gams

(1932) Die Klimatische Begrenzung von Pflanzenarealen und Die Verteilung der Hygrischen Kontinentalitaet in den Alpen. Zeitschrift derGesllschaftt FürErdkundee ZuBerlinniundefined. Verlag: Dietrich Reimer.

33.

Geburek

Hiess

Litschauer

, et al. (2012) Temporal pollen pattern in temperate trees: Expedience or fate? Oikos 121(10): 1603–1612.

34.

Gottardini

Cristofori

Cristofolini

, et al. (2004) Palynological analyses on sediments of Lake Tovel (Trentino, Italy). Studi Trentini de Scienze Naturali Acta Biologica 81(Suppl. 2): 147–154.

35.

Guo

, et al. (2020) Modern pollen and its relationship with vegetation and climate in the mu us desert and surrounding area, northern China: Implications of palaeoclimatic and palaeocological reconstruction. Palaeogeography Palaeoclimatology Palaeoecology 547: 109699.

36.

Hicks

(1994) Present and past pollen records of Lapland forests. Review of Palaeobotany and Palynology 82: 17–35.

37.

Hjelle

(1998) Herb pollen representation in surface moss samples from mown meadows and pastures in western Norway. Vegetation History and Archeobotany 7: 79–96.

38.

Hjelle

(1999) Modern pollen assemblages from mown and grazed vegetation types in western Norway. Review of Palaeobotany and Palynology 107(1-2): 55–81.

39.

Huang

Chen

(2018) Modern pollen assemblages from human-influenced vegetation in northwestern China and their relationship with vegetation and climate. Vegetation History and Archaeobotany 27(6): 767–780.

40.

International Civil Avian Organization (1993) Manual of the ICAO Standard Atmosphere (Extended to 80 Kilometres (262 500 Feet), 3rd edn. Montreal, QC: International Civil Avian Organization.

41.

Jackson

Wong

(1994) Using forest patchiness to determine pollen source areas of closed-canopy pollen assemblages. Journal of Ecology 82: 88–99.

42.

Julier

ACM

Jardine

Adu-Bredu

, et al. (2018) The modern pollen–vegetation relationships of a tropical forest–savannah mosaic landscape, Ghana, West Africa. Palynology 42(3): 324–338.

43.

Karger

Conrad

Böhner

, et al. (2017) Climatologies at high resolution for the earth’s land surface areas. Scientific Data 4: 170122.

44.

Karger

Schmatz

Dettling

, et al. (2020) High-resolution monthly precipitation and temperature time series from 2006 to 2100. Scientific Data 7: 248.

45.

Kasper

Leuschner

Walentowski

, et al. (2022) Winners and losers of climate warming: Declining growth in Fagus and Tilia vs. stable growth in three Quercus species in the natural beech–oak forest ecotone (western Romania). Forest Ecology and Management 506: 119892.

46.

Leuschner

(2020) Drought response of European beech ( Fagus sylvatica L.)—A review. Perspectives in Plant Ecology Evolution and Systematics 47: 125576.

47.

LIDAR PAT (2014) Portale Cartografico Trentino. Available at: https://siat.provincia.tn.it/stem/

48.

Lisitsyna

Hicks

(2014) Estimation of pollen deposition time-span in moss polsters with the aid of annual pollen accumulation values from pollen traps. Grana 53(3): 232–248.

49.

Markgraf

(1980) Pollen dispersal in a mountain area. Grana 19: 127–146.

50.

Mitchell

Jones

(2005) An improved method of constructing a database of monthly climate observations and associated high-resolution grids. International Journal of Climatology 25: 693–712.

51.

Moore

Webb

Collison

(1991) Pollen Analysis. Oxford: Blackwell Scientific publications.

52.

Morales-Molino

Devaux

Georget

, et al. (2020) Modern pollen representation of the vegetation of the Tagus Basin (central Iberian Peninsula). Review of Palaeobotany and Palynology 276: 104193.

53.

Noce

Caporaso

Santini

(2020) A new global dataset of bioclimatic indicators. Scientific Data 7: 398.

54.

Odasso

Miori

Gandolfo

(2018) I tipi forestali del Trentino: descrizione e aspetti dinamici. Servizio foreste e fauna, Provincia autonoma di Trento.

55.

Oksanen

Blanchet

Kindt

, et al. (2016) Vegan: Community ecology package. R package version 2.3-5. Available at: www.cran.r-project.org/web/packages/vegan (accessed 27 January 2022).

56.

Ortu

Klotz

Brugiapaglia

, et al. (2010) Elevation-induced variations of pollen assemblages in the north-western alps: An analysis of their value as temperature indicators. Comptes Rendus Biologies 333(11-12): 825–835.

57.

Papadopoulou

Tsiripidis

Panajiotidis

, et al. (2022) Testing the potential of pollen assemblages to capture composition, diversity and ecological gradients of surrounding vegetation in two biogeographical regions of southeastern Europe. Vegetation History and Archaeobotany 31(1): 1–15.

58.

Pedrotti

Venanzoni

(1987) La vegetazione di un bacino glacio-carsico sull'altipiano di Folgaria (Trentino). Webbia 41(2): 289–304.

59.

Pidek

Svitavská -Svobodová

van der Knaap

, et al. (2013) Pollen percentage thresholds of Abies alba based on 13-year annual records of pollen deposition in modified Tauber traps: Perspectives of application to fossil situations. Review of Palaeobotany and Palynology 195: 26–36.

60.

Pignatti

(2013) Forests of broadleaved trees and shrubs at low elevation. In: Pignatti

Pignatti

(eds) Plant Life of the Dolomites: Vegetation Structure and Ecology. Berlin, Heidelberg: Springer, pp.71–119.

61.

Poska

Pidek

(2010) Pollen dispersal and deposition characteristics of Abies alba, Fagus sylvatica and Pinus sylvestris, Roztocze region (SE Poland). Vegetation History and Archaeobotany 19(2): 91–101.

62.

Provincia Autonoma di Trento (2011) Il castagno e il noce nella Provincia di Trento. Prov. Autonoma di Trento, Assessorato all’Agricoltura, Foreste, Turismo e Promozione. Artimedia Casa Editrice, Trento. Fifth Edition, p.116.

63.

Punt

Blackmore

Clarke

GCS

, et al. (1976–2009) The Northwest European Pollen Flora. Vol. I-IX. Amsterdam: Elsevier.

64.

Quamar

Tiwari

Thakur

(2021) The modern pollen–vegetation relationship in Jammu, India: A comparative appraisal. Acta Palaeobotanica 61(1): 1–19.

65.

Räsänen

Suutari

Nielsen

(2007) A step further towards quantitative reconstruction of past vegetation in Fennoscandian boreal forests: Pollen productivity estimates for six dominant taxa. Review of Palaeobotany and Palynology 146(1–4): 208–220.

66.

R Core Team (2021) R: A Language and Environment For Statistical Computing. Vienna, Austria: R Foundation for Statistical Computing. Available at: www.R-project.org/ (accessed 27 January 2022).

67.

Reille

(1992–1998) Pollen Et Spores d’Europe Et d’Afrique Du Nord. Marseille: Laboratoire de Botanique historique et palynologie.

68.

Rodrigo-Comino

Salvia

Quaranta

, et al. (2021) Climate aridity and the geographical shift of olive trees in a Mediterranean Northern Region. Climate 9: 64.

69.

Sartori

Mancabelli

(2009) Carta dei suoli del Trentino. Scala 1:250000. Consiglio per la Ricerca e la Sperimentazione in agricoltura (CRA). Esperia srl, Laves.

70.

Schüler

Hemp

Behling

(2014) Relationship between vegetation and modern pollen-rain along an elevational gradient on Kilimanjaro, Tanzania. The Holocene 24(6): 702–713.

71.

Senn

Tinner

Felde

, et al. (2022) Modern pollen – Vegetation – Plant diversity relationships across large environmental gradients in northern Greece. The Holocene 32(3): 159–173.

72.

Servera-Vives

Mus Amezquita

Snitker

, et al. (2022) Modern analogs for understanding pollen-vegetation dynamics in a Mediterranean mosaic landscape (Balearic Islands, western Mediterranean). The Holocene 32: 716–734.

73.

Sjögren

van der Knaap

Huusko

, et al. (2008a) Pollen productivity, dispersal, and correction factors for major tree taxa in the Swiss alps based on pollen-trap results. Review of Palaeobotany and Palynology 152: 200–210.

74.

Sjögren

van der Knaap

O. Kaplan

, et al. (2008b) A pilot study on pollen representation of mountain valley vegetation in the central Alps. Review of Palaeobotany and Palynology 149: 208–218.

75.

Soepboer

Sugita

Lotter

, et al. (2007) Pollen productivity estimates for quantitative reconstruction of vegetation cover on the Swiss Plateau. The Holocene 17(1): 65–77.

76.

Sugita

(1994) Pollen representation of vegetation in Quaternary sediments: Theory and method in patchy vegetation. Journal of Ecology 82: 881–897.

77.

Tabares

Mapani

Blaum

, et al. (2018) Composition and diversity of vegetation and pollen spectra along gradients of grazing intensity and precipitation in southern Africa. Review of Palaeobotany and Palynology 253: 88–100.

78.

Urrego

Silman

Correa-Metrio

, et al. (2011) Pollen-vegetation relationships along steep climatic gradients in western Amazonia. Journal of Vegetation Science 22(5): 795–806.

79.

Vallè

Furlanetto

Maggi

, et al. (2019) Concepts and methodology to quantitatively reconstruct climate from pollen data. Geografia Fisica e Dinamica Quaternaria 42(2): 225–234.

80.

von Humboldt

Bonpland

(1807) Essay on the Geography of Plants. Reprinted in 2010. Chicago, IL: University of Chicago Press.

81.

Zhang

Steiner

(2022) Projected climate-driven changes in pollen emission season length and magnitude over the continental United States. Nature Communications 13: 1234.

82.

Zhang

Duo

Pang

, et al. (2018) Modern pollen assemblages and their relationships to vegetation and climate in the Lhasa Valley, Tibetan Plateau, China. Quaternary International 467: 210–221.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

2.09 MB