AMQ32(2) 117-130 Furlanetto et al.pub ECOLOGICAL CLIMATOLOGY ALONG AN ELEVATIONAL TRANSECT IN THE OUTER BELT OF THE ITALIAN ALPS: MODERN POLLEN, VEGETATION AND CLIMATE   Giulia Furlanetto1,2, Lorena Garozzo2, Michele Brunetti3, Cesare Ravazzi2 1 Univ. of Milano - Bicocca, Dept. of Environmental and Earth Sciences, Milano, Italy. 2 CNR - Institute for the Dynamics of Environmental Processes (IDPA), Laboratory of Palynology and Palaeoecology, Milano, Italy. 3 CNR- Institute of Atmospheric Sciences and Climate (ISAC), Bologna, Italy. Corresponding author: G. Furlanetto ABSTRACT: Montane vegetation is traditionally known to be particularly sensitive to climate changes. The strong elevational cli- matic gradient that characterizes mountain areas results in a steep ecological slope, with several ecotones occurring in a small area. Modern pollen deposition is significantly predicted by both vegetation cover and pollen production; in turn, each predictor is significantly predicted by elevation and climate. Analyses of modern pollen deposition are essential to calibrate fossil pollen se- quences accounting for site topography and depositional process, and thus for pollen-based palaeoenvironment and palaeocli- mate reconstructions. This study analyzes the relationships among modern pollen assemblages, vegetation and climate along an elevational gradient in the outer belt of the European Alps. Results of Canonical Correspondence Analysis (CCA) demonstrated a general good agreement with previous studies, which identified elevation as the main gradient in the variation of modern pollen and vegetation assemblages in elevational transects. Modern pollen assemblages have been studied in pollen traps and moss samples from different vegetation communities along an elevational transect (stretching from 1240 to 2390 m asl), as well as the vegetation using the Braun-Blanquet system up to the 10 m radius scale, field vegetation surveys and aerial photographs for plant cover of the main species for larger surfaces. Moss samples are assumed to record an average of several years of pollen deposi- tion and can be profitably used as analogues for fossil pollen assemblages; while pollen traps can be expressed as Pollen Accu- mulation Rates (PAR) and used as a modern reference to estimate past plant population densities. Alnus viridis, the main woody species forming dwarf forests in the oceanic-type timberline ecotone, shows a specific elevational PAR arrangement under mod- ern climate conditions. Strong pollen producers (e.g. Pinus sylvestris/mugo, Picea, Castanea, Corylus and Ostrya) display en- hanced uphill wind-transport to subalpine and alpine zones leading to wider pollen belts with less defined boundaries than vegeta- tion. To overcome these limitations, potential indicator pollen taxa of alpine/subalpine belts (Vaccinium, Rhododendron, Loiseleu- ria) documented in this study and PAR of timberline species (e.g. Alnus viridis) could be useful. Thus, if it is possible to identify the major vegetation types and ecotones by means of their modern pollen deposition (e.g. timberline ecotone), other limits, poorly marked by changes in pollen dispersal (e.g. treeline) are not resolved by pollen proxies alone. Keywords Alps, elevational transects, modern pollen assemblages, vegetation, climate. Available online http://amq.aiqua.it ISSN (print): 2279-7327, ISSN (online): 2279-7335 Alpine and Mediterranean Quaternary, 32 (2), 2019, 117 - 130 1. INTRODUCTION Ecological climatology is an interdisciplinary frame- work to understand the functioning of terrestrial ecosys- tems in the climate system (Bonan, 2016). Relationships between climate, ecosystems and forcing factors may be effectively understood by setting up training sets along ecogradients, and using proxies that allow to syn- thesize spatial variables such as vegetation (pollen pro- portions), terrestrial biomass (pollen accumulation rates), fire impact (number of fire events in the last 100 years), etc. In turn, analyses of modern training sets are es- sential for a better understanding of fossil palaeoecolog- ical records in a particular region, and thus for pollen- based palaeoenvironment and palaeoclimate recon- structions. The variability associated to modern pollen deposition sampled in an extended Alpine calibration set responds mostly to July temperature (Furlanetto et al., 2019), a climate variable strictly related to elevation. Annual precipitation was instead found as the most sig- nificant climate variable for the Armentarga fossil record covering the last 10.3 ka cal BP (Furlanetto et al., 2018). Nowadays annual precipitation is not significantly related to elevation as July temperature but is instead strongly controlled by regional climate (Furlanetto et al., 2018). Very few studies of modern pollen deposition have been attempted in mountain environments of southern European countries (Court-Picon et al., 2005, 2006; Cañellas-Boltà et al., 2009; Ortu et al., 2010). Moss samples are commonly used as surface samples for local modern pollen deposition. They are assumed to record an average of several years of pollen deposition (Räsänen et al., 2004; Pardoe et al., 2010; Lisitsyna & https://doi.org/10.26382/AMQ.2019.08 118 Furlanetto G. et al. Hicks, 2014) and can be profitably used as analogues for fossil pollen assemblages. Instead, pollen traps can be expressed both as percentages and Pollen Accumu- lation Rates (PAR). PAR measurements can be used as a modern reference to estimate past plant population densities buffering the study site (Hicks, 2001; Tinner & Theurillat, 2003; Badino et al., 2018; Furlanetto et al., 2018). Modern references allow also the so-called “site calibration” to be defined, accounting for site-specific features affecting either climate parameters and pollen deposition, such as local topography and sedimentary processes, and the local influence of runoff and, most conspicuously, snow runoff on PAR, barely considered in traditional palaeoecological studies. The aim of this study is to investigate modern pol- len assemblages-vegetation-climate relationships along an elevational gradient. This paper presents a new ele- vational transect developed in the Upper Brembana Valley (Orobie Alps), recording a typical mid-latitude ecoclimatic elevational gradient under oceanic cold tem- perate climate (Furlanetto et al., 2018). Here modern pollen assemblages, vegetation and climate have been collected at 16 sampling sites placed along an eleva- tional gradient stretching from 1240 m asl to 2390 m asl. The transect has been designed (1) to estimate the relationships linking modern assemblages, vegetation, elevation and climate; (2) to understand how is the mod- ern vegetation reflected in modern pollen assemblages from pollen traps and moss samples; (3) to calculate specific elevational PAR arrangement under modern climate conditions; (4) to provide a robust modern refer- ence for reliable palaeoenvironmental and quantitative interpretations of past vertical shifts and changes in vegetation composition, as recognized in the pollen records from high-elevation sites. 2. STUDY AREA The Upper Brembana Valley elevational transect extends along the uppermost section of a typical preal- pine valley open to moist southern winds orographically forced uphill, between 1100 m and the headwalls at 2500-2700 m asl (Fig. 1). This region is marked by a Dfc climate accordingly to Köppen-Geiger classification (Peel et al., 2007), i.e. nival, persistently wet throughout the year; locally, the annual temperature is 1.3 °C and July lapse rate ranges from 0.56 °C/100 m in the moun- tain belt to 0.77 °C/100 m above the open forest limit (Furlanetto et al., 2019). Annual precipitation is not sig- nificantly related to elevation (Tab. 1). The site actually represents a snowfall extreme in the entire Alps, snow accumulation being concentrated in late winter and spring, and net snow accumulation summing up to 7 m at 1950 m asl (1964-1973 period, Belloni & Pelfini, 1990). The glacier Equilibrium Line Elevation under cur- rent climate conditions exceeds the mountain peaks; it was about 2600 m in 1994 AD (Caccianiga et al., 1994). The vegetation of the area provides a representative example of elevational ecological gradient in an oceanic climatic regime common in the Eastern and Northern Alps, heavily affected by pastoralism in the last 4 ka, which however, did not alter the fundamental traits of the timberline structure (Furlanetto et al., 2018, Fig. 2). 3. METHODS: DEVELOPMENT OF AN ELEVATIONAL TRANSECT AND SAMPLING STRATEGY A transect composed by 16 sampling sites was developed (Fig. 1). The precise location and descriptive information of the sampling sites are listed in Tab. 1. At the center of each site, a pollen trap was located on the ground fixed to iron bars, in order to avoid snow runoff at the soil level. Three moss samples were collected within Fig. 1 - Sampling sites for modern pollen assemblages and vegetation data stretching along a typical mid-latitude ecoclimatic elevational gradient under oceanic cold temperate climate in the Upper Brembana Valley. For this study, only the upper elevational reach, spanning a lapse of 1200 m from 1200 to 2400 m asl, was considered. 119 Ecological climatology along an elevational transect a circular area of 1.8 m radius (Fig. 3, zone A) and mixed together (Cañellas-Boltà et al., 2009), while con- centric circles of 1.8 and 10 m-radius were considered for vegetation surveys following the Braun-Blanquet method (Fig. 3, zone A and B). Furthermore, main ar- boreal species were listed and a cover estimation was carried out within a 100 m radius circle with the help of orthoimages (Fig. 3, zone C). 3.1. Canonical Correspondence Analysis (CCA) Ordination techniques are considered appropriate for assessing and displaying the relationship between pollen percentages recorded in modern pollen assem- blages, vegetation data, elevation and climatic parame- ters. These statistical analyses have been frequently used to detect structures or patterns within pollen and vegetation data. They provide an efficient low- dimensional representation of complex multivariate data and serve to indicate their major gradients and the rela- tive contribution of each taxon to each of these gradi- ents. All the plants present an ecological optimum and tolerance to most environmental gradients, justifying the use of a biological unimodal response model (ter Braak, 1987, ter Braak and Prentice, 1988). In this work, Ca- nonical Correspondence Analysis (CCA), the con- strained form of Correspondence Analysis (CA) (ter Braak 1986) was used as direct gradient analysis to highlight the relationships between pollen, vegetation, elevation and climate. Several CCA ordinations have been performed using elevation, July temperature, Janu- Tab. 1 - Site metadata for the Upper Brembana Valley elevational transect. ary temperature and annual precipitation as the sole constraining variable to assess their importance on the variance of pollen and vegetation (1.8 m and 10 m radi- us) assemblages (Tab. 2). Prior to the CCA analysis, pollen percentages were square-root transformed, for variance stabilization (Prentice, 1980) and taxa with percentages below 2%, were removed. The significance of all CCA were tested with ANOVA (p <0.05, after 999 unrestricted permutations). 3.2. Estimation of climate normals along the eleva- tional transect Site-specific instrumental temperature and precipi- tation series, covering the 1951-2015 period, were re- constructed for each sampling site by means of the anomaly method (New et al., 2000; Mitchell & Jones, 2005) as described in Brunetti et al. (2012), to obtain a climate reconstruction as much representative of the specific locations as possible, thanks to the huge amount of meteorological observations available for Italy and surrounding areas. The closest meteorological sta- tion (Foppolo, 1520 m asl) is situated ca 4 km far from site L 1 and ca 9.5 km far from site L 24. The interpola- tion procedure consists in the independent reconstruc- tion of the climatological normals over a given reference period (i.e. climatologies) and the departures from them (i.e. the anomalies): the former, characterized by re- markable spatial gradients, were reconstructed by evalu- ating the local dependence of the meteorological varia- ble on elevation (Brunetti et al., 2014; Crespi et al., 2018), and a high spatial density network of stations (even if with a limited temporal coverage) was used; anomalies are linked to climate variability and are char- acterized by a higher spatial coherence, therefore, a more simple interpolation technique and a low spatial density of stations are enough, but long temporal cover- age and accurate homogenization (i.e. the procedure that removes the non-climatic signals introduced by stations and instruments relocation, changes in meas- urement practices and so on) are mandatory. Finally, climatologies and anomalies were superimposed to get temperature and precipitation monthly series in absolute values for each sampling site. Information about the techniques and their accuracy are provided in Brunetti et al. (2012, 2014) and Crespi et al. (2018). From these monthly series, mean temperature of the warmest month (July), coldest month (January) and the total annual precipitation over the 1981-2010 period were calculated for the sampled sites (Tab. 1). 120 Fig. 2 - Picture showing the structure of the oceanic-type timberline ecotone at the headwall of the Brembana Valley in the outer Italian Alps. Elevation values refer to current limits in the studied area (in brackets the maximum elevations recorded in the Orobie Alps). The natural extent of forest vegetation is the result of long-term effects of pastoralism, wildfire, and climate changes. According to paleobotani- cal data, the current timberline structure still reflects the dominance of pristine dwarf forests which become established in the second half of the Holocene (Furlanetto et al., 2018). Furlanetto G. et al. 121 Tab. 2 - Proportions of variance explained by each explanatory variable as the sole constrained and permutational-repeated measures analysis (ANOVA) obtained for modern pollen assemblages (pollen traps and moss samples) and vegetation data (surveys at 1.8 m and 10 m radius circles). Fig. 3 - Site sampling method scheme used for vegetation survey. Ecological climatology along an elevational transect 3.3. Vegetation data: surveying strategy (Braun- Blanquet system) At each sampling site, vegetation surveys were carried out using the Braun-Blanquet abundance/ dominance scale (Braun-Blanquet, 1979) within circular areas of 1.8 m radius (ca 10 m2) and 10 m radius (ca 314 m2). In this system, a semi-quantitative abundance- dominance index is used to estimate the plant cover for each species, according to the following scale: 5 (80- 100%), 4 (60-80%), 3 (40-60%), 2 (20-40%), 1 (1-20%), + (scarce), r (rare). Finally, within the circular areas of 100 m radius was estimated the cover of the main spe- cies. This methodology was focused on producing good estimates of the abundance of the major taxa (the pro- ducers of the pollen types which are the major compo- nents of current and past pollen deposition in the study area) rather than recording every species present. Com- pared to plant species identification, pollen usually pre- sents lower taxonomic resolution, in most cases only reaching genus or family level. To facilitate comparative data analyses plant species recorded in the vegetation surveys were grouped into higher taxonomic levels to match their corresponding pollen types and presented as percentages. 3.4. Pollen analysis - modern pollen assemblages At the center of each site, a pollen trap was locat- ed (Fig. 3, zone A). The trap has a circular opening on top (4.5 cm in diameter), through which the airborne pollen can freely enter. Glycerine and thymol were placed in the trap to inhibit the growth of mould and the opening was covered by a 2 mm mesh fixed to the trap with duct tape. Pollen monitoring data represents the period 2016-2017. In each site, three moss polsters were also collected within a circular area of 1.8 m radius (ca 10 m2) and mixed into one sample. Mosses were sampled down to, but not including, the mineral soil. Samples were processed following standard methods at the Laboratory of Palynology and Palaeoecology of CNR-IDPA in Milan, after adding Lycopodium tablets for pollen and pollen-slide microcharcoal concentration estimations (Stockmarr, 1971). A minimum count of 600 pollen grains was obtained at each sampling site, aquat- ics, spores and non-pollen palynomorphs were excluded from the pollen sum. Identification was carried out at x400, x630 and x1000 magnifications under a Leica DM -LB light microscope. Pollen identification followed Moore et al. (1991), Punt & Blackmore (1976-2009), Reille (1992-1998), Beug (2004) and the CNR pollen reference collection. Microbiological particles were named after van Geel et al. (1981), van Geel (1978) and van Geel & Aptroot (2006) and expressed as percent- ages out of the pollen sum. Pollen-slide microcharcoal particles were counted under a light microscope at 400x. Black, completely opaque and angular fragments (Clark, 1988) were identified as charcoal and grouped in two size classes (10-50 μm and 50-250 μm length). 3.5. Pollen Accumulation Rates (PAR) Pollen monitoring using annual pollen traps helps to quantify and calibrate pollen deposition with present- day vegetation and meteorological conditions, as well as providing insights into local pollen dispersal and sedi- mentation mechanisms (Hicks, 2001). Pollen Accumula- tion Rates (PAR) measurements can also be used as a modern reference to estimate past plant population den- sities buffering the study site (Hicks, 2001; Tinner & Theurillat, 2003; Badino et al., 2018). Modern PAR val- ues were calculated for Alnus viridis, one of the main timberline ecotone shrub species in the outer oceanic districts of Central Alps. 4. RESULTS 4.1. Canonical Correspondence Analysis (CCA) This study aims at emphasizing pollen/vegetation/ climate relationships at a local scale. Indeed, although extra-local and regional non-arboreal pollen is present in the modern pollen spectra, herbaceous pollen assem- blages appear to reflect predominantly local vegetation rather than wider landscape units, as also demonstrated by Caseldine & Pardoe (1994), Hjelle (1999) and Pardoe (2001). The variances explained by elevation and each climate parameter (Tjuly, Tjan and Pann), calculated with CCA ordinations with unique constrained, are pre- sented in Tab. 2. The Canonical Correspondence Analy- sis (CCA) of the vegetation data and pollen assemblag- es suggests that the variation in vegetation and pollen is strongly correlated to the elevational gradient (Tab. 2). In both vegetation data and pollen assemblages eleva- tion explains more variance than the climate parame- ters, although alike percentages of variance are ex- plained by Tjuly and Tjan. For the pollen traps the vari- ance explained by elevation is 17.0%, 16.6% by Tjuly, 16.9% by Tjan and 12.7% by Pann. For the moss sam- ples the variance explained by elevation is 14.6%, 14.4% by Tjuly, 14.5% by Tjan and 10.9% by Pann. Pann is not significant in the moss samples assemblage (p > 0.05, after 999 unrestricted permutations). For the vegetation 1.8 m the variance explained by elevation is 12.6%, 12.3% by Tjuly, 12.4% by Tjan and 12.0% by Pann. Finally for the vegetation 10 m the variance ex- plained by elevation is 15.6%, 15.4% by Tjuly, 15.4% by Tjan and 14.0% by Pann. 4.2. Modern pollen assemblages and their relation- ship to vegetation The pollen spectra of 18 pollen traps and 16 moss samples are presented in synthetic percentage dia- grams, which include selected pollen types and total microcharcoal concentrations (Figs. 4-5). Plant species recorded in the 10 m vegetation surveys, grouped into higher taxonomic levels to match their corresponding pollen types are presented in a synthetic percentage diagram (Fig. 6). In the site L 1 located at 1244 m asl (Fig. 6) the main vegetation species are Corylus avellana, Picea abies and Larix decidua among trees and shrubs; Festu- ca sp. and Laserpitium halleri and Laserpitium krapfii among herbs. In the pollen traps the dominant taxa are Picea, Corylus and Poaceae (Fig. 4, TL1-TL2) while Corylus, Ostrya type, Fraxinus excelsior type, Pinus sylvestris/mugo, and Larix represent the main taxa in the moss samples (Fig. 5, ML1-ML2). Pollen taxa from broad-leaved tree (e.g. Castanea, and Ostrya type) oc- cur in high proportions (Figs. 4a-5a). Alnus viridis reach- 122 Furlanetto G. et al. dra (10%), while the site L 10 shows the 50% of Agrostis sp., 30% of Festuca scabriculmis and 30% cover of San- guisorba dodecandra. The sites L 2 and L 5 are respec- tively located in mowed leans surrounded by thickets of broad-leaved trees and in a pasture at the border of a mixed coniferous forest. In the 10 m vegetation survey (L 2) Pimpinella major has the 30% of cover, followed by 10% of Alchemilla vulgaris, Geranium sylvaticum, Hera- es high cover percentages (10-30%) in sites L 3 and L 4 located next to streams and also in the site L 10 (30%) located in an avalanche corridor/stream (Fig. 6), this abundance in Alnus viridis is also presented in modern pollen assemblages (Figs. 4a-5a). Sites L 3 and L 4 show also high cover percentages of Festuca scabricul- mis (50-70%), Galium lucidum (30%), Pimpinella major (10%) and the endemic tall herb Sanguisorba dodecan- 123 Fig. 4 - Percent diagram obtained from pollen traps from the Upper Brembana Valley elevational transect showing selected pollen taxa and main vegetation zones. Ecological climatology along an elevational transect cleum sphonodylium, Rumex scutatus, Trifolium pre- tense and Polygonum bistorta; in the site L 5 Laburnum alpinum has 50% of cover, Picea abies 20%, Larix de- cidua 10%, among herbs Festuca pratensis has 20% of cover. The site L 7, a grassland along an avalanche corridor presents 10% of Picea abies and Larix decidua (Fig 6a); among herbs 30% of Calamagrostis sp., Rhinanthus sp. and Sanguisorba dodecandra. This lat- ter herb is well represented in the pollen trap (trap TL7, Fig. 4b) reaching 18%. The sites L 12-L 18 are located in the upper subalpine belt with Ericaceae moorlands (Erica carnea, Vaccinium gaultherioides, Vaccinium myrtillus, Rhododendron ferrugineum and Loiseleuria 124 procumbens) and Alnus viridis thickets with Larix decid- ua individuals; larch treeline is located at 2220 m. Juni- perus nana shows high cover percentages in the sites L 12 and L 15, that are little reflected in the pollen percent- ages. In the sites (L 12-L 18) the pollen trap and moss sample assemblages are dominated by Alnus viridis, Pinus sylvestris/mugo, and Poaceae, the Ericaceae moorland is well represented by Vaccinium while Erica, Rhododendron and Loiseleuria show sporadic occur- rences (Figs. 4a-5a). Above 2100 m asl the wind-driven uphill transport of pollen from the lower belts shows increased percentages of Pinus sylvestis/mugo, Betula, Corylus, Ostrya type, Fagus, and Carpinus betulus in Fig. 5 - Percent diagram obtained from moss samples from the Upper Brembana Valley elevational transect showing selected pollen taxa and main vegetation zones. Furlanetto G. et al. the moss samples and increased percentages of Picea, Betula, Corylus and Castanea in the pollen traps. In sites (L 21-L 24) the 10 m vegetation is dominated by Festuca scabriculmis, Festuca quadriflora, Carex sem- pervirens (L 21); Salix herbacea, Chrysanthemum al- pinum and Potentilla aurea (L22); Cirsium spinosissi- mum, Poa alpina, Rumex alpestris and Potentilla aurea (L 24). The site L 24 (2390 m asl) is located in an exten- sive pasture, reflected in modern pollen assemblages by high percentages of Poaceae and of dung spores (Sordariaceae undiff., Sporormiella-like spores, and Podospora). 4.3. Pollen Accumulation Rate (PAR) Alnus viridis, the main woody species forming ex- tensive dwarf forests in the Upper Brembana Valley timberline ecotone (Fig. 2), shows a specific elevational PAR arrangement under modern climate conditions (Tab. 3). The highest PAR values in the local elevational transect are observed in sites L3 and L10 (1420 and 1940 m asl, respectively) in river corridors and waterfalls with mountain alder scrub. Here PAR values range be- tween 2400 and 3100 pollen grains cm-2 yr-1 (Tab. 3, traps TL3 and TL10). On the other hand, pollen traps at subalpine elevation (Tab. 3, traps TL16, TL18; respec- tively 2100 and 2180 m asl), located at the border of mountain alder scrub with continuous understory of Eri- caceae, display PAR values of about 700 pollen grains cm-2 yr-1. 125 Fig. 6 - Percent diagram of plant cover of selected plant species from the Upper Brembana Valley elevational transect recorded in the 10 m vegetation surveys. Species were grouped into higher taxonomic levels to match their corresponding pollen types. In order to allow com- parison with pollen assemblages the same taxa presented in the pollen diagrams were selected. Ecological climatology along an elevational transect 5. DISCUSSION: POLLEN-VEGETATION-ELEVATION AND CLIMATE RELATIONSHIPS Our data are in agreement with previous studies which identified elevation as the main gradient control- ling modern pollen assemblages and vegetation distri- bution across mountain regions (Bonnefille et al., 1993; Vincens et al., 1997; Weng et al., 2004; Court-Picon 2005, 2006; Rull, 2006; Cañellas-Boltà et al., 2009). The Canonical Correspondence Analysis (CCA) showed that elevation explains the major percentage of variance in both vegetation data and pollen assemblages (Tab. 2); although alike percentages of variance are explained by the temperature climate parameters (Tjuly and Tjan). All the temperature parameters vary accordingly with eleva- tion due to a stable lapse rate. July lapse rate ranges from 0.56 °C/100 m in the mountain belt to 0.77 °C/100 m above the open forest limit (Furlanetto et al., 2019). A lower percentage of variance is explained by annual precipitation (Tab. 2), that is not significantly related to elevation as temperature but is instead strongly con- trolled by regional climate (Fig. 2B in Furlanetto et al., 2018). In spite of the uphill pollen transport, pollen- elevation and climate relationships are strong since pollen is directly correlated to elevation and climate. Pollen and vegetation data are independent. 5.1. Uphill pollen transport The uphill transport of pollen by wind to subalpine and alpine zones was first recorded in the European Alps (Markgraf, 1980; David, 1997; Frei, 1997; Bru- giapaglia et al., 1998; Ortu, 2002). Markgraf (1980) showed that, in a steep valley-slope system of Alpine type, the proportion of pollen dispersal from local uphill winds in daytime decreases with increasing elevation in favor of regional (medium-distance) dispersal. This high- elevation dispersal component relies significantly on pollen washout by rain, which cleans the atmosphere of suspended pollen (Markgraf, 1980). Pollen belts tend to be wider and with less defined boundaries than vegeta- tion; this is likely due to the homogenizing effect of wind dispersal, especially from lower to upper elevational levels. Thus, if it is possible to identify the major vegeta- tion types by means of their modern pollen deposition, it becomes more difficult to identify specific thresholds (e.g. treeline). The over-representation of Pinus in the pollen as- semblages is well documented by numerous authors (e.g. Broström et al. 1998; Brugiapaglia et al. 1998; Ortu et al. 2005, 2006; Court- Picon et al., 2005, 2006, Cañellas-Boltà et al., 2009). The strong pollen signal of Pinus sylvestris/mugo is recorded in the moss samples above the treeline (Fig. 5a, ML17-24), while Pinus mugo is not recorded in the vegetation data (Fig. 6). In the study area, isolated stands of mountain dwarf pine (Pinus mugo) form part of the timberline structure at 1800-2000 m asl at dry edaphic sites. A dark coniferous forest (Picea abies, Abies alba, Larix decidua) occurs only on shaded slopes, therefore Picea abies appears in the 10 m vegetation surveys only in sites L 1, L 5 and L 7 (Fig. 6a). In the pollen traps and moss samples Picea shows a marked uphill transport to subalpine and alpine zones in the pollen traps (Fig. 4a). Deciduous trees of the mountain belt (e.g. Corylus, Betula, Ostrya, Castanea) also show uphill wind transport. Castanea uphill transport is marked in pollen traps (Figs. 4a) while Corylus and Ostrya show high percentages in the moss samples ML16-21 between 2106 and 2341 m asl (Fig. 5a). The influence of upslope wind dispersal of medium- to long-distance transported pollen types (e.g. Castanea and Ostrya type) have to be taken into account and interpreted with caution in the fossil records. 5.2. Potential indicator taxa of subalpine/alpine belts Extensive mountain alder shrublands develop in the subalpine elevational belt under oceanic, cool tem- perate climates, that hinder cembran pine fitness. Edaphic geodiversity allows mountain alder to compete with dwarf pine (Pinus mugo), the latter also tolerating oceanic climates, but escaping edaphic wetness. Alnus viridis vegetation needs a continuous water supply dur- ing the growing season, enhanced by high annual snow- fall rate and good soil water balance (Richard, 1968, 1969; Mauri and Caudullo, 2016). In Alpine oceanic 126 Tab. 3 - Modern Alnus viridis Pollen Accumulation Rate values based on pollen monitoring data in the Upper Brembana Valley, representing the period 2016-2017. The location of the sites is shown in Fig. 1. Furlanetto G. et al. climates, its ecological range expands downstream es- pecially along avalanche corridors but also on northern slopes, due to a general increase in soil moisture, and thus species optima are shifted towards lower eleva- tions (e.g. Fig. 2); for this reason, it shows lower eleva- tion optima compared to continental inner valleys (Furlanetto et al., 2019). Alnus viridis PAR measure- ments from pollen traps at subalpine elevation (2100- 2180 m asl), located at the border of mountain alder dwarf forest can be used as a modern reference to esti- mate past plant population densities (Furlanetto et al., 2018). Caution is due because the pattern of decreasing pollen production with elevation (Furlanetto et al., 2018) also affects PAR values. Pollen Productivity Estimates (PPEs) were not calculated from modern pollen samples because regional (104-105 km2—REVEALS model) and local (<1 km2 up to 5 km2—LOVE model) vegetation reconstruction methods were not applied to the fossil record (Furlanetto et al., 2018) as the application of LOVE model (fossil pollen from small study sites) was not possible. The LOVE model estimates the vegetation composition within the source area subtracting the back- ground pollen from the pollen signal. As pollen records from large sites closed to the study area for regional vegetation composition were lacking this model was not applied to the fossil record. Ericaceae is considered an excellent indicator of the montane/subalpine belts (Cañellas-Boltà et al., 2009). In our study, the identification of pollen types within this family allows to highlight different elevational patterns of pollen and relative parent taxa. Vaccinium, Rhododendron and Loiseleuria show a consistent pollen -vegetation elevational pattern restricted to the timber- line belt between 2000 and 2300 m asl (Fig. 7). Where- as Erica pollen type shows a wider distribution especial- ly in pollen traps from 1422 to 2345 m asl (Fig. 4a) even though it is recovered in the 10 m vegetation surveys between 1550 and 2000 m asl (Fig. 6). Also Juniperus shows an elevational pattern in both vegetation and pollen traps with major abundance at the lower limit of the open forest, ca. 2000 m (Fig. 7). 6. CONCLUSIONS Relationships between climate, ecosystems and forcing factors may be effectively understood by setting up training sets along ecogradients, and using proxies that allow to synthesize spatial variables such as vege- tation (pollen proportions) and terrestrial biomass (pollen accumulation rates). The development of an elevational transect in the Upper Brembana Valley (Central Italian Alps) represents a first step for analyzing ecological 127 Fig. 7 - Histograms showing the elevational distribution of plant cover % (10 m vegetation surveys) and of pollen % (moss samples and pollen traps) for potential indicator taxa of subalpine/alpine belts. Ecological climatology along an elevational transect gradients with multivariate analysis of modern pollen assemblages as synthetic proxies addressing relation- ships between modern vegetation, elevation and cli- mate. The results of CCA analysis demonstrated a gen- eral good agreement with previous studies, which identi- fied elevation as the main gradient in the variation of modern pollen and vegetation assemblages in mountain areas); although alike percentages of variance are ex- plained by the temperature climate parameters (Tjuly and Tjan). Alnus viridis, the main woody species forming the dwarf forests in the timberline ecotone, shows a specific elevational PAR arrangement under modern climate conditions. The highest PAR values are observed in sites from river corridors and waterfalls with mountain alder scrub. The uphill transport of pollen by wind to subalpine and alpine zones implies wider pollen belts with less defined boundaries than vegetation. This is mainly due to the increase in extra-local transport of strong pollen producers (e.g. Pinus sylvestris/mugo, Picea, Castanea, Corylus and Ostrya) and lead to some homogenization of pollen assemblages. To overcome these limitations, potential indicator pollen taxa of alpine/subalpine belts (Vaccinium, Rhododendron, Loiseleuria) documented in this study and PAR of timberline species (e.g. Alnus viridis) could be useful. Thus, if it is possible to identify the major vegetation types and ecotones by means of their modern pollen deposition (e.g. timberline ecotone), other limits, poorly marked by changes in pollen disper- sal (e.g. treeline) are not resolved by pollen proxies alone. ACKNOWLEDGMENTS This paper is the result of Giulia Furlanetto’s PhD research project sponsored by the Milano-Bicocca Uni- versity, Dept. of Earth and Environmental Sciences, with the co-tutoring and support of the CNR-IDPA. This pa- per is a contribution to the CNR-IDPA Research Project DTA.AD001.112 "Quaternary environmental history, ecological changes and mutual interactions with human societies". Suggestions provided by Séverine Fau- quette, Donatella Magri and Elda Russo-Ermolli im- proved the quality of the manuscript. Special thanks go to Migliorini Enzo, manager of the Rifugio Longo and to CAI for logistic support; to Renata Perego and Massimo Domenico Novellino for support during the field work. This research was supported by a grant from Parco Regionale delle Orobie Bergamasche (Delib. 12/2017). REFERENCES Badino F., Ravazzi C., Vallè F., Pini R., Aceti E., Brunet- ti M., Champvillair E., Maggi V., Maspero F., Pere- go R., Orombelli G. (2018) - 8800 years of high- altitude vegetation and climate history at the Rutor Glacier forefield, Italian Alps. Evidence of middle Holocene timberline rise and glacier contraction. Quaternary Science Reviews, 185, 41-68. Belloni S., Pelfini M. (1990) - La nevosità in Lombardia nel periodo 1964-1973. Natura Bresciana, 25, 63- 108. Beug H.J. (2004) -. Leitfaden der Pollenbestimmung für Mitteleuropa und angrenzende Gebiete. Verlag Dr. Friedrich Pfeil, München, Germany. Bonan G. (2016). Ecological Climatology. Concepts and applications. Cambridge University Press. Bonnefille R., Buchet G., Friis I.B., Kelbessa E., Moham- med M.U. (1993) - Modern pollen rain on an altitu- dinal range of forests and woodlands in South West Ethiopia. Opera Botanica, 121, 71-84. Braun-Blanquet J. (1979) - Fitosociologica. Blume, Ma- drid. Broström A., Gaillard M.J., Ihse M., Odgaard B. (1998) - Pollen-landscape relationships in modern ana- logues of ancient cultural landscapes in southern Sweden - a first step towards quantification of veg- etation openness in the past. Vegetation History and Archaeobotany, 7(4), 189-201. Brugiapaglia E., de Beaulieu J.L., Guiot J., Reille M. (1998) - Transect de pluie pollinique et étagement de la végétation dans le massif du Taillefer (Isère, France) - Géographie physique et Quaternaire, 52 (2), 209-218. Brunetti M., Lentini G., Maugeri M., Nanni T., Simolo C., Spinoni J. (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. Brunetti M., Maugeri M., Nanni T., Simolo C., Spinoni J. (2014) - High-resolution temperature climatology for Italy: interpolation method intercomparison. International Journal of Climatology, 34(4), 1278- 1296. Caccianiga M., Ravazzi C., Zubiani P. (1994) - Storia del ghiacciaio del Trobio (Alpi Orobie, Bergamo) e colonizzazione della vegetazione nelle aree libera- te dopo la Piccola Età Glaciale. Natura Bresciana, 29, 65-96. Cañellas-Boltà N., Rull V., Vigo J., Mercadé A. (2009) - Modern pollen-vegetation relationships along an altitudinal transect in the central Pyrenees (southwestern Europe). The Holocene, 19(8), 1185 -1200. Caseldine C., Pardoe H. (1994) - Surface pollen studies from alpine/sub-alpine southern Norway: applica- tions to Holocene data. Review of Palaeobotany and Palynology, 82(1), 1-15. Clark J.S. (1988) - Particle motion and the theory of charcoal analysis: source area, transport, deposi- tion, and sampling. Quaternary Research, 30(1), 67-80. Court-Picon M., Buttler A., 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(1), 13-39. Court-Picon M., Buttler A., de Beaulieu J.L. (2006) - Modern pollen/vegetation/land-use relationships in mountain environments: an example from the Champsaur valley (French Alps). Vegetation histo- ry and Archaeobotany, 15(3), 151-168. Crespi A., Brunetti M., Lentini G., Maugeri M. (2018) - 1961-1990 high-resolution monthly precipitation 128 Furlanetto G. et al. climatologies for Italy. International Journal of Cli- matology, 38, 878-895. David F. (1997) - Holocene tree limit history in the north- ern French Alps stomata and pollen evidence. Review of Paleobotany and Palynology, 97, 227- 237. Frei T. (1997) - Pollen distribution at high elevation in Switzerland: Evidence for medium range transport. Grana, 36, 34-38. Furlanetto G., Ravazzi C., Pini R., Vallè F., Brunetti M., Comolli R., Novellino M.D., Garozzo L., Maggi V. (2018) - Holocene vegetation history and quantita- tive 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. Furlanetto G., Ravazzi C., Badino F., Brunetti M., Champvillair E., Maggi V. (2019) - Elevational transects of modern pollen samples - site specific temperatures as a tool for paleoclimate recon- structions in the Alps. The Holocene, 29(2), 271- 286. Hicks S. (2001) - The use of annual arboreal pollen deposition values for delimiting tree-lines in the landscape and exploring models of pollen disper- sal. Review of Palaeobotany and Palynology, 117 (1-3), 1-29. Hjelle K.L. (1999) - Modern pollen assemblages from mown and grazed vegetation types in western Norway. Review of Palaeobotany and Palynology, 107(1), 55-81. Lisitsyna O.V., Hicks S. (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. Markgraf V. (1980) - Pollen Dispersal in a Mountain Area. Grana, 19, 127-146. Mauri A., Caudullo G. (2016) - Alnus viridis in Europe: distribution, habitat, usage and threats. In: San- Miguel-Ayanz J, De Rigo D, Caudullo G, Houston Durrant T, Mauri A (eds) European Atlas of Forest Tree Species. Publication Office of the European Union, Luxembourg, p. 68. Mitchell T.D., Jones P.D. (2005) - An improved method of constructing a database of monthly climate ob- servations and associated high-resolution grids. International Journal of Climatology, 25(6), 693- 712. Moore P.D., Webb J.A., Collinson M.E. (1991) - Pollen analysis. Blackwell Scientific Publications, Oxford. New M., Hulme M., Jones P. (2000) - Representing twentieth-century space-time climate variability. Part II: Development of 1901-96 monthly grids of terrestrial surface climate. Journal of Climate, 13, 2217-2238. Ortu E. (2002) - Reconstruction sur base pollinique de la dynamique de la végétation tardiglaciaire et holo- cène dans les Alpes Maritimes Italiennes. PhD Thesis, University d'Aix-Marseille, III, F. Ortu E., David F., Caramiello R. (2005) - Effet des para- mètres locaux sur le développement de la végéta- tion dans le Vallon de St. Anna di Vinadio (Alpes Maritimes; Italie) - Écoscience, 12, 122-135. Ortu E., Brewer S., Peyron O. (2006) - Pollen-inferred palaeoclimate reconstructions in mountain areas: problems and perspectives. Journal of Quaternary Science, 21(6), 615-627. Ortu E., Klotz S., Brugiapaglia E., Caramiello R., Sinis- calco C. (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), 825-835. Pardoe H.S. (2001) - The representation of taxa in sur- face pollen spectra on alpine and sub-alpine glaci- er forelands in southern Norway. Review of Palae- obotany and Palynology, 117(1), 63-78. Pardoe H.S., Giesecke T., van der Knaap W.O., Svi- tavská-Svobodová H., Kvavadze E.V., Panajiotidis S., Gerasimidis A., Pidek I.A., Zimny M., Święta- Musznicka J., Latałowa M., Noryśkiewicz A.M., Bozilova E., Tonkov S., Filipova-Marinova M.V., van Leeuwen J.F.N., Kalniņa L. (2010) - Compar- ing pollen spectra from modified Tauber traps and moss samples: examples from a selection of woodlands across Europe. Vegetation History and Archaeobotany, 19, 271-283. Peel M.C., Finlayson B.L., McMahon T.A. (2007) - Up- dated world map of the Köppen-Geiger climate classification. Hydrology and Earth System Scienc- es, 11(5), 1633-1644. Prentice I.C. (1980) - Multidimensional scaling as a re- search tool in Quaternary palynology: a review of theory and methods. Review of Palaeobotany and Palynology, 31, 71-104. Punt W., Blackmore S. (Eds) (1976-2009) - The North- west European Pollen Flora. Vol. I-IX. Elsevier Publishing Company. Räsänen S., Hicks S., Odgaard B.V. (2004) - Pollen deposition in mosses and in a modified ‘Tauber trap’ from Hailuoto, Finland: what exactly do the mosses record? Review of Palaeobotany and Pal- ynology, 129, 103-116. Reille M. (1992-1998) - Pollen et spores d’Europe et d’Afrique du Nord. Laboratoire de Botanique histo- rique et palynologie, Marseille. Richard L. (1968) - Écologie de l'aune vert (Alnus viri- dis): facteurs climatiques et édaphiques. Docu- ments pour la Carte de Végétation des Alpes, 6, 107-158. Richard L. (1969) - Une interprétation éco-physiologique de la répartition de l'aune vert (Alnus viridis). Do- cuments pour la Carte de Végétation des Alpes, 7, 7-23. Rull V. (2006) - A high mountain pollen-altitude calibra- tion set for palaeoclimatic use in the tropical An- des. The Holocene, 16, 105-107. Stockmarr J. (1971) - Tablets with spores used in abso- lute pollen analysis. Pollen et Spores, 13, 615-621. ter Braak C.J. (1986) - Canonical correspondence analy- sis: a new eigenvector technique for multivariate direct gradient analysis. Ecology, 67(5), 1167- 1179. ter Braak C.J. (1987) - The analysis of vegetation- environment relationships by canonical corre- spondence analysis, in: Theory and models in vegetation science. Springer Netherlands, pp. 69- 129 Ecological climatology along an elevational transect 77. ter Braak C.J., Prentice I.C. (1988) - A theory of gradient analysis. Advances in Ecological Research, 18, 271-317. Tinner W., Theurillat J.P. (2003) - Uppermost limit, ex- tent, and fluctuations of the timberline and treeline ecocline in the Swiss Central Alps during the past 11,500 years. Arctic, Antarctic, and Alpine Re- search, 35(2), 158-169. van Geel B. (1978) - A palaeoecological study of Holo- cene peat bog sections in Germany and The Neth- erlands, based on the analysis of pollen, spores and macro- and microremains of fungi, algae, cormophytes and animals. Review of Palaeobota- ny and Palynology, 25, 1-120. van Geel B., Bohncke S.J.P., Dee H. (1981) - A palaeo- ecological study of an upper Late Glacial and Hol- ocene sequence from 'De Borchert', The Nether- lands. Review of Palaeobotany and Palynology, 31, 367-449. van Geel B., Aptroot A. (2006) - Fossil ascomycetes in Quaternary deposits. Nova Hedwigia 82, 313-329. Vincens A., Ssemmanda I., Roux M., Jolly D. (1997) - Study of the modern pollen rain in Western Ugan- da with a numerical approach. Review of Palaeo- botany and Palynology, 96(1), 145-168. Weng C., Bush M.B., Silman M.R. (2004) - An analysis of modern pollen rain on an elevational gradient in southern Peru. Journal of Tropical Ecology, 20 (01), 113-124. 130 Furlanetto G. et al. Ms. received: February 22, 2019 Final text received: May 10, 2019