Determining the long-term changes in biodiversity and provisioning services along a transect from Central Europe to the Mediterranean

Abstract

Climate, land use and fire are strong determinants of plant diversity, potentially resulting in local extinctions, including rare endemic and economically valuable species. While climate and land use are decisive for vegetation composition and thus the species pool, fire disturbance can lead to landscape fragmentation, affecting the provisioning of important ecosystem services such as timber and raw natural resources. We use multi-proxy palaeoecological data with high taxonomic and temporal resolution across an environmental gradient to assess the long-term impact of major climate shifts, land use and fire disturbance on past vegetation openness and plant diversity (evenness and richness). Evenness of taxa is inferred by calculating the probability of interspecific encounter (PIE) of pollen and spores and species richness by palynological richness (PRI). To account for evenness distortions of PRI, we developed a new palaeodiversity measure, which is evenness-detrended palynological richness (DE-PRI). Reconstructed species richness increases from north to south regardless of time, mirroring the biodiversity increase across the gradient from temperate deciduous to subtropical evergreen vegetation. Climatic changes after the end of the last ice age contributed to biodiversity dynamics, usually by promoting species richness and evenness in response to warming. The data reveal that the promotion of diverse open-land ecosystems increased when human disturbance became determinant, while forests became less diverse. Our results imply that the today’s biodiversity has been shaped by anthropogenic forcing over the millennia. Future management strategies aiming at a successful conservation of biodiversity should therefore consider the millennia-lasting role of anthropogenic fire and human activities.

Keywords

climate ecosystem services evenness-detrended palynological richness fire land use plant diversity

Introduction

Concerns about future biosphere shifts in response to global change include loss of biodiversity, possibly through unprecedented disappearance of species which may comprise mass extinctions (Mooney and Mace, 2009; Overpeck et al., 2003; Pereira et al., 2010; Pimm et al., 1995; Thuiller et al., 2005). The two main drivers of global change are climatic shifts in response to anthropogenic greenhouse gas forcing and land use comprising economic megatrends, such as globalization and its effects on the environments. Coupled with globalization are fire-regime shifts. In the last decades, the role of wildfire in controlling both supporting services (e.g. net primary productivity and nutrient cycling) and provisioning (e.g. timber production, crop yields) has received little attention (e.g. Millennium Ecosystem Assessment UE, 2005). The strong impact of fire on vegetation (e.g. in southern Europe) and thus on one of the most important ecosystem components emphasizes the relevance of this scientific gap. Fire not only affects important ecosystem properties like plant biomass, nutrient cycling and soil stability, but it also determines species composition at short annual (e.g. Moreno and Oechel, 1991; Pausas et al., 1999) to long centennial timescales (Bennett et al., 1992; Delarze et al., 1992; Tinner et al., 1999), for instance, by promoting fire-resistant pioneer species at the expense of sensitive late-successional ones (Colombaroli et al., 2007; Tinner et al., 2000). Besides its cultural and aesthetic value, biodiversity is a key component of ecosystems, maintaining ecosystem functioning over time, such as biomass production (Cardinale et al., 2012; Isbell et al., 2011; Marquard et al., 2009; Reich et al., 2012).

In this study, we review the palaeoenvironmental evidences of climate change, land use and fire impact on biodiversity and vegetation openness, that is, open grassland versus forest ecosystems, the latter providing important provisioning services for societies (e.g. through firewood, food and building materials). We selected key-sites that are distributed along a marked environmental gradient from Central Europe to the Mediterranean, one of the most diverse regions of the world, with over 25,000 native plant species (Medail and Quezel, 1999; Myers et al., 2000). The Mediterranean is also a cultural diversity hotspot, given that many mentifacts (including writing, sciences, engineering, law and medicine) that are now common in the Western World originated from the Mediterranean region and were spread over Europe, including the British Isles by the Greco-Roman Civilization. Palaeorecords offer the unique opportunity to explore the linkages between natural and cultural diversity over millennia (e.g. Willis and Birks, 2006). Our key-sites are representative for European beech (Lobsigensee, Swiss Plateau), mixed fir-oak (Lago di Origlio, southern Pre-Alps), deciduous and evergreen oak (Massaciuccoli, Tuscany), evergreen oak–olive–mastic forests (Gorgo Basso, Sicily) and the corresponding human derivatives (e.g. grasslands, heaths, shrublands, orchards, maquis and garrigues). Our case studies have good-to-excellent chronological control, they are located in biogeographical key positions, and they document important vegetational and environmental changes following climate variability, human disturbance and fire. We use pollen-based reconstructions of past biodiversity changes and available pollen and charcoal records to discuss: (1) how climate and fire affected biodiversity during the Lateglacial/Holocene transition (Lobsigensee and Lago di Origlio) and during the early and mid-Holocene (Lago di Massaciuccoli, Gorgo Basso), (2) how farming and anthropogenic-driven fire regimes altered biodiversity and forest fragmentation, and finally (3) whether this information can be used as a historical baseline for forest management in ecosystems with high cultural heritage values, such as the sweet chestnut forests in southern Switzerland and Italy or the Mediterranean maquis. Long-term data focusing on responses of ecosystems to changes in climate, fire and human impact are particularly useful where the maintenance and protection of naturalness is a management goal, such as in Natural Parks and natural reserves.

Materials and methods

Study sites

The sites are located across a N–S transect spanning from northern Switzerland to Southern Italy (Figure 1). Lobsigensee (47°02′N, 7°18′E, 514 m a.s.l.) is a small (8 ha) lake located in the Swiss Plateau (northern Switzerland). Climate is typically temperate central European, with mild summers (18°C average), cool winters (−1°C January mean) and annual precipitation around 900 mm. Today’s vegetation in the area is highly impacted by agriculture, but sparse relict forests of Picea abies and Fagus sylvatica are present. Lago di Origlio (46°03′N, 8°56′E, 416 m a.s.l.) is a 8 ha lake in Ticino (southern Pre-Alps, Switzerland). Compared with sites north of the Alps, the region has higher summer temperature (mean July c. 22°C), mild winters (c. 3°C January mean) and high annual rainfall (up to 1800 mm). The most common species in the sub-Mediterranean vegetation is the sweet chestnut tree (Castanea sativa); small relict mixed-oak stands comprise, for example, Quercus petraea, Quercus pubescens, Ulmus, Tilia, Fraxinus and Acer. Lago di Massaciuccoli (43°50′N, 10°20′E, 0 m a.s.l.) is a large (700 ha) coastal lake located in north Tuscany (Italy). The climate is typically Mediterranean with dry summers, in contrast to the rainy summers in the Alps. Summers are warm (c. 22°C July mean), winters mild (c. 7°C) and annual precipitation is c. 900 mm. The vegetation belongs to the meso-Mediterranean belt, with mixed broadleaved forests consisting of deciduous Quercus robur, Q. pubescens, Carpinus betulus, C. sativa and evergreen Quercus ilex, Pistacia lentiscus, Rhamnus alaternus and Phillyrea angustifolia. Finally, Gorgo Basso (37°37′N, 12°39′E, 6 m a.s.l.) is a small (3 ha) lake located in the thermo-Mediterranean belt of southern Sicily (Figure 1). Summers are hot (c. 28°C July), winters very mild (c. 12°C January temperature) and annual precipitation around 500 mm. Most precipitation at this site falls during the winter. Around the site, relicts of evergreen broadleaved Mediterranean forest occur (Q. ilex, Q. coccifera and P. lentiscus) together with cultivated fields, vineyards and orchards (e.g. Olea europaea). Further information about the sub-, meso-, and thermo-Mediterranean vegetation belts is provided, for example, in Lang (1994).

Pollen, charcoal data and chronology

Pollen, microscopic charcoal and chronology are available from previous studies. Pollen and spore data reflect the degree of vegetation openness (AP% vs NAP%, respectively arboreal and non-arboreal pollen percentages) and land-use indicators (e.g. pollen of crops and weeds), while climate variability is assessed by referring to pollen-independent palaeoclimatic reconstructions from the study area (e.g. chironomids, diatoms, lake levels). At all sites considered, only terrestrial plant material was used for radiocarbon dating; microscopic charcoal was estimated using the same approach as for pollen.

For Lobsigensee, original pollen and microscopic charcoal data were published in Ammann (1989) and Tinner et al. (2005). The chronology is based on 51 accelerator mass spectrometry (AMS)-radiocarbon dates from plant macrofossils and bulk material (Ammann, 1989; van der Knaap and Ammann, 1997); however, we only considered terrestrial plant macrofossils, following Tinner et al. (2005). For Lago di Origlio, pollen, microscopic charcoal data and the chronology (seven ²¹⁰Pb and 25 AMS-radiocarbon dates) come from Tinner et al. (1998, 1999). The pollen, microscopic charcoal and chronology for Lago di Massaciuccoli (7 AMS-radiocarbon dates) and Gorgo Basso (10 AMS-radiocarbon dates) are from Colombaroli et al. (2007) and Tinner et al. (2009), respectively. For the calculation of the pollen sum, aquatic plants, spores and non-pollen palynomorphs (NPPs) were excluded at all sites. Multiproxy evidence such as plant macrofossils, chironomids, cladocera and isotopes on ostracodes and molluscs are available for the Swiss Plateau (Ammann, 1989; Ammann et al., 2000), chironomids for the southern Pre-Alps (Samartin et al., 2012), diatoms for coastal Tuscany (Marchetto et al., 2008) and lake-level changes for southern Sicily (Magny et al., 2012).

Diversity indexes

We used different pollen-based diversity measures, to consider as many diversity aspects as possible. First, we used rarefaction analysis to calculate palynological richness (PRI), which is a proxy for species richness at local to landscape scale (Birks, 2007; Birks and Line, 1992) and has been applied in many palaeoecological studies in both northwestern European (Birks and Line, 1992; Giesecke et al., 2013; Lotter, 1999; Odgaard, 1999; Seppä, 1998) and Mediterranean environments (Bisculm et al., 2012; Colombaroli et al., 2007, 2008; Tinner et al., 1999). Rarefaction analysis estimates the number of pollen taxa that would be encountered if the pollen sum was to be kept constant (Birks and Line, 1992). In our case, we used a constant pollen sum for each site, which was standardized on the minimum pollen sum (Gorgo Basso, n = 207), across all sites.

Second, we estimated the palynological evenness as probability of interspecific encounter (PIE; see Hurlbert, 1971), to understand the extent to which the total number of taxa in a sample is influenced by few dominant species (van der Knaap, 2009). For instance, sample with pollen taxa from high-pollen producers (such as Pinus sylvestris) may result in fewer taxa detected in the pollen sample, and thus in an underestimation of the total number of taxa in the landscape. This effect led some authors to conclude that PRI is a mixed proxy for both species richness and evenness (Odgaard, 2006). To account for this possible bias, we also calculated an evenness-detrended palynological richness (DE-PRI). This detrending mainly resulted in higher values of PRI in samples with lower evenness and vice versa. We applied the detrending by using the relationship between evenness and richness, as estimated by an ordinary least squares (OLS) linear regression equation for each site (Supplementary material S1), with the dependent variable (y) being PRI and the independent variable (x) being palynological evenness. Since OLS-equations may underestimate the slope of the linear relationship between the variables, when variables are measured with errors (e.g. Cornbleet and Gochman, 1979), we also tested other regression models, including major axis (MA) and standard major axis (SMA) (see Legendre, 2001; Legendre and Legendre, 1988; Supplementary material S2). The obtained values from the OLS equation (palynological evenness scaled on PRI values) were then subtracted to the original expected number of taxa (according to PRI). The obtained residuals of pollen richness (PRI−palynological evenness) were then distributed around the pollen richness average (residuals + average pollen richness). Only if both PRI and DE-PRI agree, we assume that our species richness estimates are unaffected by evenness effects. For all the analysis, we used the program R statistics (R Development Core Team, 2008).

Results and interpretations

Spatial and climatic trends of palynologically inferred plant diversity (species richness, evenness)

The reconstruction of past biodiversity changes from pollen data has been extensively used in palaeoecological studies across Europe (e.g. Bennett et al., 1992; Birks, 2007; Colombaroli et al., 2013; Giesecke et al., 2013; Lotter, 1999). Although its interpretation requires some care (e.g. Odgaard, 1999, 2006; Weng et al., 2006), it is a valuable measure of long-term biodiversity changes. Difficulties comprise pollen production, dispersal and deposition, which may bias the linkage between plant and pollen diversity. For instance, low pollen producers might be underestimated, while high producers (e.g. Pinus, Betula) may produce assemblages with artificially low richness values. To avoid such evenness effects, we interpret our data considering both PRI and DE-PRI (see methods).

Our reconstructed trends (Figure 2) show that changes in richness (PRI) and evenness (PIE-index) mainly occur during periods of climatic changes (Lateglacial to Holocene transition) or increasing land use (Neolithic, Bronze and Iron Age). In general, PRI and DE-PRI values are highest (30–50) in the thermo-Mediterranean vegetation (Gorgo Basso), and lowest (10–20) in the temperate vegetation (Lobsigensee), with intermediate values (c. 20–30) in the sub-Mediterranean and meso-Mediterranean belts (Lago di Origlio and Lago di Massaciuccoli; see Figure 2). This suggests that increasing plant diversity from Central Europe to the Mediterranean is well mirrored in PRI.

The northernmost record (Lobsigensee, north of Alps) shows a sharp decrease in PRI and palynological evenness at the transition from the Oldest Dryas to the Bølling/Allerød interstadial (c. 14,600; Figure 2), when temperatures increased rapidly by about c. 3–4°C (Lotter et al., 2012; Rasmussen et al., 2006). A similar PRI decline at 14,600 cal. BP also occurs at Lago di Origlio (southern Pre-Alps; Figure 2). During the Bølling/Allerød, the PIE-index differs markedly among the two sites (high vs low PIE-index for Lobsigensee and Lago di Origlio, respectively; Figure 2). The predominance of few taxa (i.e. low PIE-index) around Lago di Origlio may have directly affected PRI, leading to an underestimation of plant species richness. We solve the potential evenness distortions of PRI by considering the DE-PRI, which suggest higher species richness values around Lago di Origlio during the Bølling/Allerød (higher DE-PRI; Figure 2).

The transition to the Holocene (11,600 cal. BP) is marked by slight increases (Lobsigensee) or no changes in PRI (Lago di Origlio), suggesting different effects of climatic warming on plant diversity at the two sites. However, increasing palynological evenness is recorded at the Lateglacial/Holocene transition (c. 11,600 cal. BP) at both sites, pointing to less predominant species in response to rapid climatic warming (Ammann et al., 2000; Schwander et al., 2000). Accordingly, DE-PRI declines (Lobsigensee) or remains relatively stable (Lago di Origlio), possibly showing that our species richness estimates are slightly affected (i.e. overestimated) by evenness effects.

A marked change in PRI occurs during the Neolithic at both sites (c. 7500–4200 cal. BP), suggesting that land use (archaeologically documented onset of farming) significantly contributed to increased species richness north and south of the Alps. The good agreement between PRI and DE-PRI suggests that our richness estimates are robust and unaffected by evenness effects. PRI and DE-PRI increase again together at both sites during the late Bronze Age (c. 3000 cal. BP; Figure 2), pointing to diverse open landscape following intensification of land use (Tinner et al., 2003). From the Bronze Age until modern times, PRI and DE-PRI increased gradually at both sites, indicating that the introduction of new agropastoral techniques and/or crops (e.g. C. sativa plantations during Roman times) further increased species richness on the Swiss Plateau and at the foot of the southern Pre-Alps. Low PRI and DE-PRI values at Lago di Origlio since the 1950s are instead explained by declining land-use activities and spontaneous re-afforestation (Tinner et al., 1998), confirming the positive link between agropastoral activities and biodiversity since the onset of the Neolithic.

Further south along the Mediterranean coast our PIE-index, PRI and DE-PRI records cover the past 10,000 (Gorgo Basso) to 7000 (Massaciuccoli) years. At Lago di Massaciuccoli, in the Tuscan coastlands, PRI and DE-PRI are highly variable during the phase following the end of the Versilian sea transgression (c. 8000–7000 cal. BP; see Blanc, 1942; Lambeck et al., 2004). This suggests oscillating biodiversity dynamics in response to highly disturbed environments, as indicated by the deposition of several metres of sand in the area (Colombaroli et al., 2007). Subsequently, PRI and DE-PRI increase steadily after c. 6000 cal. BP, with the exception of a period between 5500 and 5000 cal. BP, when lake levels were low, likely because of drier conditions (Marchetto et al., 2008). This increase in species richness by c. 6000 cal BP is in good agreement with the development at Lago di Origlio and Lobsigensee (Figure 2). At Gorgo Basso, high PIE, PRI and DE-PRI values before 7000 cal. BP suggest diverse vegetation during the early Holocene and at the beginning of the mid-Holocene. A decline in PIE, PRI and DE-PRI occurs at c. 7000 cal. yr BP, during the early Neolithic. Archaeological data indeed document that people left the coast and moved inland during the early Neolithic (Leighton, 1999), possibly causing an impoverishment of vegetation through land abandonment. However, as for Lago di Massaciuccoli, Lago di Origlio and Lobsigensee, PRI and DE-PRI increased markedly during the past 3000 years, that is, since the Late Bronze Age (c. 3000 cal. BP; Figure 2). On the basis of combined archaeological and archaeobotanical evidence (see Ammann, 1989; Colombaroli et al., 2007; Tinner et al., 1999, 2009), we interpret this spatio-temporal trend as a significant increase of biodiversity in response to agricultural activities during the past 3000 years (Figures 2 and 3).

The link between species richness and vegetation as reconstructed from pollen

The PRI and DE-PRI inferred changes of species richness as well as the pollen-inferred vegetation evenness can be explained by the vegetation composition, as reconstructed from the pollen assemblages. At our northernmost site Lobsigensee, the sharp decrease in species richness and evenness during the Lateglacial was caused by succession processes involving Juniperus, which was followed by Betula and Pinus (Ammann, 1989). The slight increase in evenness at the Lateglacial/Holocene transition likely reflects the gradual replacement of Pinus with deciduous Quercus sp. and Corylus.

In southern Switzerland and Northern Italy, afforestation began already at c. 18,000–17,500 cal. BP with the expansion of Juniperus shrublands, thus millennia before similar vegetational processes occurred in Central and northern Europe (e.g. Lobsigensee). However, forested environments appeared only after c. 16,000 cal. BP, when open Juniperus shrublands were replaced by stone-pine forests (Pinus cembra) at Origlio, in Canton Ticino and in neighbouring areas in Northern Italy (Tinner et al., 1999; Vescovi et al., 2007). This vegetation shift probably occurred in response to abrupt climatic warming after the end of Heinrich event 1 (Samartin et al., 2012). Afforestation processes 17,500–14,500 cal. BP, including the expansion of stone-pine forests did not affect evenness and species richness significantly, as suggested by our palaeodiversity measures (PIE-index, PRI and DE-PRI). Instead, evenness was strongly reduced when Scotch-pine (P. sylvestris) forests replaced stone-pine forests at the onset of the Bølling/Allerød interstadial at c. 14,700 cal. BP. The combined interpretation of PRI and DE-PRI suggests that species richness did not decline as dramatically as evenness in the Lateglacial forests. Evenness increased again at around 13,000–12,500 cal. BP, when thermophilous taxa such as Quercus, Ulmus and Fraxinus expanded around Origlio and elsewhere in southern Switzerland and Northern Italy (Tinner et al., 1999, 2005; Vescovi et al., 2007), most likely in response to climatic warming (Samartin et al., 2012). High evenness may mean a rather diverse forest (see PRI and DE-PRI; Figure 2), with a lowered predominance of a few different species, in the newly established nemoral forests of the region.

After the onset of the Holocene, diversity remained constant until the early Neolithic (c. 7500 cal. BP) north and south of the Alps, when a significant reduction in Abies alba, Tilia and other broadleaved taxa occurred (Figure 4). Subsequently, species richness and evenness increased to reach high values during the Neolithic (between 5500 and 5000 cal. BP), following intensification of land use as shown by anthropogenic indicators (e.g. Cerealia-t; Figure 3). The further increase of PRI observed at both sites of Lobsigensee and Lago di Origlio, which started at c. 3000 cal. BP (Figure 2), was due to an expansion of plants of open habitats, for example, cultivated fields and meadows (e.g. Cerealia-t, Plantago lanceolata, Rumex acetosa-t, Artemisia). An expansion of herbaceous species and an increase of palynologically inferred species richness and evenness following land-use intensification has been shown by continuous high-resolution time-series analyses from the Swiss Rhone Valley lowlands (Colombaroli et al., 2013).

A close link between land use and increasing species richness also occurred at the Mediterranean sites. At Lago di Massaciuccoli, mixed evergreen and deciduous forests re-established after 7000 cal. BP, replacing early successional Pinus stands (Colombaroli et al., 2007). The re-established vegetation included A. alba, which today is not growing around the site, and the evergreen oak Q. ilex (Bellini et al., 2009; Colombaroli et al., 2007). Low values of PRI between 5500 and 5000 cal. BP can be instead attributed to the establishment of salty-marshland ecosystems (Chenopodium and Salicornia; Colombaroli et al., 2007) possibly in response to drier climatic conditions (Marchetto et al., 2008). Between 4000 and 2000 cal. BP, forest ecosystems at Massaciuccoli were rather closed, with Q. ilex, F. sylvatica and other thermophilous and mesophilous species. Similar to the situation in the lowlands of the Rhine (Lobsigensee) and Po-River catchment (Lago di Origlio), increasing land use led to a diversification of habitats at the Tuscanian Sea shore, with more diverse grassland from the Roman and until Mediaeval time. Indeed, the period after c. 2000 cal. BP corresponded to the establishment in the Massaciuccoli area of the high diverse modern maquis vegetation (Pistacia, Erica, Juniperus sp.). Its degraded form (garrigue) established after c. 1000–700 cal. BP, with a predominance of Phillyrea (Colombaroli et al., 2007), and led to slightly less diverse ecosystems, as shown by declining PRI and DE-PRI values (Figure 2).

At Gorgo Basso, a decrease of diversity after 7000 cal. BP was associated with the transition from rather open and diverse Pistacia ecosystems to slightly poorer evergreen O. europaea-Q. ilex forests. Evergreen broadleaved Mediterranean forest had established after the collapse of intense Neolithic coastland fig–cereal cultures (Tinner et al., 2009). However, the collapse of agriculture was not the cause but more likely the consequence of the expansion of evergreen forests in response to increasing moisture availability (Tinner et al., 2009). A marked increase of species diversity occurred between 3000 and 2000 cal. BP (Figure 2) and was coupled to the conversion of forest to open land. Forest disruption was a consequence of land-use intensification during Greek and Roman colonization of coastal Sicily (Figure 3). With the demise of evergreen broadleaved forests and the creation of open habitats for farming, biodiversity reached values higher than during the pristine Pistacia maquis phase, prior to 7000 cal. BP (Figure 2).

Species richness, biomass burning and deforestation

Natural and anthropogenic fires are a strong determinant of vegetation changes in both Mediterranean and temperate ecosystems, directly affecting forest composition, and often leading to local extinctions of fire-sensitive species (e.g. A. alba, Ulmus, Tilia or Hedera helix; Colombaroli et al., 2007; Tinner et al., 1999). We compare charcoal records to pollen data and reconstructed indices of palaeobiodiversity changes to assess the long-term influence of fire on biodiversity and landscape fragmentation. North of the Alps, regional fire activity was low during the Lateglacial and early Holocene (Figure 3) when species richness had average values. Regional fire activity markedly increased around 6000–5500 cal. BP (Figure 3) during a Neolithic settlement phase around the lake (Cortaillod culture; Tinner et al., 2005), which resulted in forest reduction (AP values from c. 90% to 70%) and a first striking increase of species richness (Figure 2). Anthropogenic indicators show that fires after 6500 cal. BP were mostly driven by humans (Figure 3). Maximum Holocene biomass burning was reached during the onset of the Iron Age at c. 2700 cal. BP and again was connected to forest reduction and a marked rise in biodiversity.

In agreement with Lobsigensee, regional fire activity increased during the Neolithic at Lago di Origlio, and was at its maximum during the Iron Age (between c. 2800 and 2200 cal. BP; Figure 3). Regional fire disturbance and the creation of open land again resulted in higher species richness, confirming the pattern observed north of the Alps. PRI time-series analyses on continuous high-resolution palynological series with 10 years resolution show that anthropogenic fire disturbance significantly reduced forest diversity during the period 7100–5100 cal. BP, while open lands were established (Figure 3 and Tinner et al., 1999). In general, diversity increased despite this loss of natural communities due to the creation of new habitats such as fields, heaths and meadows.

The same approach revealed that at Lago di Massaciuccoli, forest species (e.g. A. alba, Q. ilex) were also strongly reduced by Neolithic fire (Figures 3 and 4 and Colombaroli et al., 2007), while during the Bronze Age, species richness remained constant (Figure 3). From the Roman and until Mediaeval time, fire increased together with taxa of open ecosystems (maquis vegetation, with Phillyrea, Pistacia, Erica) and species richness. Similarly, at Gorgo Basso, the abundance of maquis was positively correlated with higher fire incidence and species richness during the early Holocene. Species richness and maquis abundance decreased during the O. europaea–Q. ilex forest phase, when fires were at their Holocene minimum (Figures 3 and 4). Anthropogenic fire disturbance again reached maximum values during the Iron Age (c. 2700 cal. BP; Figure 3), when open land replaced evergreen broadleaved forests (AP from 80% to c. 20%; Figure 3) and species richness and evenness increased.

This striking agreement in species richness, fire activity and land-use dynamics between sites in different biomes and over 1000 km of distance can only be explained by very similar controls, which might be material–culture innovations (e.g. iron tools) and climatic changes. Subtle climatic changes, especially changing moisture availability, may have influenced biodiversity and fire activity via land-use success, that is, harvest and husbandry yields, which in turn were decisive for human population densities and deforestation pulses (Tinner et al., 2009). In this sense, we assume positive feedbacks with favourable climatic conditions leading to higher agricultural yields, growing human population densities and deforestation activities by fire, to increase the cultivated land. As a consequence, biodiversity increased in the newly created open habitats over wide areas of central and southern Europe (see Tinner et al., 2003, 2009 for details on direction and timing of Holocene climatic changes).

Discussion

Long-term changes in plant diversity before landscape transformation by people during the mid- and late Holocene

Vegetational changes on the Swiss Plateau were the result of high-amplitude climatic oscillations during the Lateglacial (Lotter, 1999; Lotter et al., 1992, 2012; Tinner et al., 2005). During this period, most of the observed changes in diversity are related to forested/non-forested conditions following rapid climatic changes, rather than fire (Figures 2 and 3). For instance, the transition from species-rich grasslands and dwarf shrublands (mainly Salix and Betula nana; see Lotter, 1999) to less diverse Juniperus shrublands or parklands (14,500 cal. BP) coincided with the beginning of rapid warming at the onset of the Bølling (Ammann et al., in press). Instead, the subsequent P. sylvestris–Betula woodland phase (14,000–11,000 cal. BP, Allerød) did not lead to any remarkable changes in species richness. Generally, Lateglacial forests and steppic tundra environments were moderately diverse (PRI c. 10–15) and not significantly species poorer than pristine early and mid-Holocene forests (Figure 2). A similar pattern, with diverse tundra dwarf shrublands before the (equally diverse) pristine forests, emerges also at Lago di Origlio in southern Pre-Alps (PRI c. 15–25).

At Gorgo Basso in Sicily, early Holocene natural maquis habitats were more diverse than mid-Holocene pristine evergreen forests (PRI 30–40 vs 20–30; Figures 1 –4). However, mid-Holocene pristine forests in coastal Sicily (PRI 20–30) were more diverse than their counterparts at Lago di Origlio and Massaciuccoli (PRI c. 20) and Lobsigensee (PRI 10–15). Overall, the impact of fire before the expansion of agriculture (c. 7500 cal. BP) was more closely connected with biodiversity changes at the Mediterranean (Gorgo Basso) rather than at the sub-Mediterranean (Lago di Origlio) and Central European (Lobsigensee) sites.

The impact of anthropogenic fire and land-use intensification on grassland and forest diversity

After the onset of the Neolithic (c. 5500 cal. BC in central Europe and Northern Italy, c. 6000 cal BC in Mediterranean Italy; see Malone, 2003), slash-and-burn practices resulted in extensive forest clearances and consequent reduction of forest ecosystems (e.g. Carcaillet, 1998; Clark et al., 1989; Colombaroli et al., 2008; Gobet et al., 2003; Vanniere et al., 2008). Land use and forest opening further increased during the Iron Age (Figure 3). In our study, 20% of NAP might already reflect substantial open habitats. Indeed, the NAP/AP ratio likely underestimates the opening of forests as evidenced by calibration studies (Soepboer et al., 2007; Sugita, 2007; Sugita et al., 2006). The very strong link between the expansion of crops and weeds, the creation of open habitats and the increase of fire activity at all study sites suggests that human impact rather than climate factors was the direct driver of biodiversity increases during the mid- and late Holocene (Figures 2 –4). The human-driven spread of grasslands from the mid- to late Holocene at our sites is mirrored in the western Mediterranean (e.g. Iberian Peninsula; Carrión et al., 2010a, 2010b; Gil-Romera et al., 2010), the Balkans and the eastern Mediterranean (e.g. Colombaroli et al., 2009; Lawson et al., 2004; Turner et al., 2008; Wagner et al., 2009), pointing to an unprecedented alteration or degradation of pristine forest ecosystems, due to large-scale land-use intensification in the entire Mediterranean realm.

Intensification of agricultural practices and the use of fire to establish and maintain open land had important consequences for forest fragmentation and plant diversity. Generally, our indicators for diversity and vegetation evenness point to the establishment of more diverse grasslands and meadows following the demise of forest ecosystems (Figure 2 and 3). At Lobsigensee, fire impact during the Neolithic (c. 4500 cal. BC) was higher than today and led to important changes in vegetation, including the replacement of original deciduous forests (Fraxinus, Tilia and Ulmus), with more fire-resistant species (e.g. resprouting Corylus avellana). Statistical analyses show that the promotion of Corylus shrublands at the expense of the primeval forests was significant at several sites north and south of the Alps, including Lago di Origlio (Tinner et al., 2005). The use of fire to maintain open fields and meadows favoured agriculture practices and livestock-grazing. This practice intensified during the Bronze and Iron Ages at Lobsigensee and Lago di Origlio, as indicated by significant correlations between charcoal and cultural pollen indicators (Tinner et al., 2005). Nevertheless, the effect of ecosystem disturbances on diversity became evident only with the Iron Age (3000–2000 cal. BP), when land clearances and recurrent fires avoided resprouting species (e.g. Corylus) to recolonize open areas. Also, diversity first increased to levels significantly higher than in the displaced natural forests (Lobsigensee PRI 20–30 vs 10–15; Lago di Origlio 25–30 vs 20; see Figures 2 and 3). Extensive land-use clearances were apparently efficient in areas with climate conditions more conducive for fire (i.e. warmer and/or drier summers). Such conditions are also ideal for agriculture (Lobsigensee and Lago di Origlio), in contrast to the cool and moist situation at Soppensee (Tinner et al., 2005). In good agreement with our palaeodiversity data, high-resolution pollen and microscopic charcoal records from the lowlands of the Swiss Rhone Valley suggest that the intensification of agriculture and the establishment of a patchy landscape resulted in a general increase in grassland diversity at the Mesolithic–Neolithic transition, when farming was introduced (5500 cal. BC; Colombaroli et al., 2013). An opposite trend, with recolonization of open land by trees and a decline of biodiversity, is supported by short-term ecological studies in areas which are experiencing farmland abandonment (Antrop, 2005; Pereira et al., 2010; Stoate et al., 2009).

Further south in the meso-Mediterranean and thermo-Mediterranean region, where fires today are more common and human impact started some centuries earlier, vegetational shifts mainly consisted of the conversion of closed deciduous-evergreen (e.g. Massaciuccoli) or pure evergreen broadleaved forests (e.g. Gorgo Basso) into maquis, garrigue and orchards. Such changes caused increased biodiversity in the long-term; however, in the short- and mid-term, biodiversity declined significantly in areas where fire affected pristine, fire-sensitive ecosystems. Two examples come from our sites Lago di Massaciuccoli and Lago di Origlio, where the unique natural forest communities completely vanished forever (Colombaroli et al., 2007; Tinner et al., 1999). Indeed, species-rich lowland forests in which A. alba was associated with Q. ilex, Phillyrea, Q. pubescens, Q. petraea, Tilia, Fraxinus and Acer do not exist anymore in the Italian Peninsula today, though they were widespread before the advent of Neolithic slash-and-burn activities (Colombaroli et al., 2007; Gobet et al., 2000; Kaltenrieder et al., 2010; Tinner et al., 1999; Wehrli et al., 2007). Only a few relicts of thermo-Mediterranean evergreen broadleaved forests still exist, for example, in coastal Sicily. Somewhat unexpected, our data show that diversity in these forests was as high as in the open cultural landscapes around Lobsigensee, Lago di Origlio and Massaciuccoli (PRI c. 20–30).

Another decline of biodiversity occurred in recent times (c. 1000–700 cal. BP) when persisting burning lead to a general simplification of forest structure and a decrease in species of maquis ecosystems (Colombaroli et al., 2007). Similar to what is observed in the sub- and meso-Mediterranean vegetation, periods with highest fire impact are also generally not associated with increased biodiversity in the thermo-Mediterranean belt (e.g. at 1200 and 2700 cal. BP at Gorgo Basso in Sicily), though increasing trends in grassland biodiversity are generally associated with intermediate values of disturbance (Connell, 1978; Grime, 1974). In this ecosystem type, the presence of low disturbances tends to promote moderately species-rich evergreen forests, whereas excessive fire disturbance generally leads to a simplification and homogenization of ecosystems (Colombaroli et al., 2013).

Conclusion

Our palaeodiversity data show that human-driven disturbances in the past played an important role in the establishment and maintenance of biodiversity in both temperate and Mediterranean ecosystems, explaining today’s highly diverse local but rather homogeneous regional ecosystems (Winter et al., 2009). Over the millennia, humans had an active role in keeping ecosystems open and diverse. Rather than unintentional (e.g. Blondel, 2006), this behaviour may come from practically acquired knowledge that ecosystems such as grassland are more productive when diverse (Isbell et al., 2011; Reich et al., 2012) and that they may provide a much diversified range of provisional food support, in contrast to monocultures.

Also, our data suggest that highest biodiversity values usually occurred at intermediate disturbance levels, which have been maintained over millennia through coppicing, grazing, controlled burning and traditional landscape management (e.g. terracing; see Blondel, 2006). Overall, diversity declined when disturbance was either absent or too severe (Connell, 1978; Grime, 1974). Thus, our palaeodiversity estimates show that biodiversity has a natural (pre-Neolithic) and an anthropogenic (agrarian) component. It is not easy to reconcile these two components at a landscape scale, since during the Holocene, opposite trends were observed in anthropogenic ecosystems, with species-losses in forested ecosystems and species-gain in open lands.

This situation may change rapidly over the next decade, as a consequence of land-use changes, like land-abandonment or intensive agriculture. From a management perspective, our data suggest that the easiest way to maintain and foster biodiversity is the maintenance of a landscape mosaic in which different disturbance and land-use intensities occur, from entirely or almost absent (e.g. forest wilderness) to rather intense (e.g. maquis). These mosaics should be sufficiently large (landscape scale) and have different functions, from provisioning services (food and wood timber) to recreation and nature protection. We anticipate that allowing very low disturbance levels in protected areas such as National Parks will promote higher diversities, if pristine species-rich forest communities are able to re-establish spontaneously after fire. However, allowing moderate fire disturbance in open areas for food production will preserve important cultural landscape and vegetation types, which were artificially created by humans over the past millennia. Examples for these are C. sativa forests, Quercus suber/Q. ilex dehesas, Ostrya carpinifolia shrublands, Calluna heathlands and Mediterranean maquis and garrigues, which harbour hundreds of different varieties of domesticated animals and cultivated plants, such as Olive trees (over 400 cultivars classified; e.g. Terral et al., 2004). These cultural landscapes of today are typical for the different European cultural landscapes; they should therefore be maintained by optimizing provisional services and biodiversity maintenance efforts.

Footnotes

Acknowledgements

We thank Pim van der Knaap for providing the data through the Alpine Pollen Database program (ALPADABA), Brigitta Ammann for support and advice and two referees for comments on the manuscript. We are also grateful to Kathy Willis and Elizabeth Jeffers, who organized the PAGES Focus 4 Biodiversity Theme Workshop in 2012 on this theme.

Funding

This study was supported by the SNSF Ambizione grant PZ00P2–126573 to DC.

References

Ammann

(1989) Late-Quaternary palynology at Lobsigensee: Regional vegetation history and local lake development. Dissertationes Botanicae 137: 1–157.

Ammann

Birks

HJB

Brooks

. (2000) Quantification of biotic responses to rapid climatic changes around the Younger Dryas – A synthesis. Palaeogeography, Palaeoclimatology, Palaeoecology 159: 313–347.

Ammann

van Leeuwen

JFN

van der Knaap

. (2012) Vegetation responses to rapid warming and to minor climatic fluctuations during the Late-Glacial Interstadial (GI-1) at Gerzensee (Switzerland). Palaeogeography, Palaeoclimatology, Palaeoecology

Antrop

(2005) Why landscapes of the past are important for the future. Landscape and Urban Planning 70: 21–34.

Bellini

Mariotti-Lippi

Montanari

(2009) The Holocene landscape history of the NW Italian coasts. The Holocene 19: 1161–1172.

Bennett

Boreham

Sharp

. (1992) Holocene history of environment, vegetation and human settlement on Catta ness, Lunnasting, Shetland. Journal of Ecology 80: 241–273.

Birks

HJB

(2007) Estimating the amount of compositional change in late-Quaternary pollen-stratigraphical data. Vegetation History and Archaeobotany 16: 197–202.

Birks

HJB

Line

(1992) The use of rarefaction analysis for estimating palynological richness from Quaternary pollen-analytical data. The Holocene 2: 1–10.

Bisculm

Colombaroli

Vescovi

. (2012) Holocene vegetation and fire dynamics in the supra-Mediterranean belt of the Nebrodi Mountains (Sicily, Italy). Journal of Quaternary Science 27: 687–698.

10.

Blanc

(1942) Variazioni climatiche ed oscillazioni della linea di riva nel Mediterraneo centrale durante l’Era Glaciale. Geologie der Meere und Binnengenwasser 5: 50–90.

11.

Blondel

(2006) The ‘Design’ of Mediterranean landscapes: A millennial story of humans and ecological systems during the historic period. Human Ecology 34: 713–729.

12.

Carcaillet

(1998) A spatially precise study of Holocene fire history, climate and human impact within the Maurienne valley, North French Alps. Journal of Ecology 86: 384–396.

13.

Cardinale

Duffy

Gonzalez

. (2012) Biodiversity loss and its impact on humanity. Nature 486: 59–67.

14.

Carrión

Fernandez

Gonzalez-Samperiz

. (2010a) Expected trends and surprises in the Lateglacial and Holocene vegetation history of the Iberian Peninsula and Balearic Islands. Review of Palaeobotany and Palynology 162: 458–475.

15.

Carrión

Fernandez

Jimenez-Moreno

. (2010b) The historical origins of aridity and vegetation degradation in southeastern Spain. Journal of Arid Environments 74: 731–736.

16.

Clark

Merkt

Muller

(1989) Post-glacial fire, vegetation, and human history on the northern alpine forelands, south-western Germany. Journal of Ecology 77: 897–905.

17.

Colombaroli

Beckmann

van der Knaap

. (2013) Changes in biodiversity and vegetation composition in the central Swiss Alps during the transition from pristine forest to first farming. Diversity and Distributions 19(2): 106–243.

18.

Colombaroli

Marchetto

Tinner

(2007) Long-term interactions between Mediterranean climate, vegetation and fire regime at Lago di Massaciuccoli (Tuscany, Italy). Journal of Ecology 95: 755–770.

19.

Colombaroli

Tinner

van Leeuwen

. (2009) Response of broadleaved evergreen Mediterranean forest vegetation to fire disturbance during the Holocene: Insights from the peri-Adriatic region. Journal of Biogeography 36: 314–326.

20.

Colombaroli

Vanniere

Emmanuel

. (2008) Fire–Vegetation interactions during the Mesolithic–Neolithic transition at Lago dell’Accesa, Tuscany, Italy. The Holocene 18: 679–692.

21.

Connell

(1978) Diversity in tropical forests and coral reefs. Science 199 (4335): 1302–1310.

22.

Cornbleet

Gochman

(1979) Incorrect least-squares regression coefficients in method-comparison analysis. Clinical Chemistry 25(3): 432–438.

23.

Delarze

Caldelari

Hainard

(1992) Effects of fire on forest dynamics in southern Switzerland. Journal of Vegetation Science 3: 55–60.

24.

Giesecke

Wolters

Jahns

. (2013) Exploring Holocene changes in palynological richness in northern Europe – Did postglacial immigration matter? PLoS One 7(12): e51624. DOI: 10.1371/journal.pone.0051624.

25.

Gil-Romera

Carrión

Pausas

. (2010) Holocene fire activity and vegetation response in South-Eastern Iberia. Quaternary Science Reviews 29: 1082–1092.

26.

Gobet

Tinner

Hochuli

. (2003) Middle to Late Holocene vegetation history of the Upper Engadine (Swiss Alps): The role of man and fire. Vegetation History and Archaeobotany 12: 143–163.

27.

Gobet

Tinner

Hubschmid

. (2000) Influence of human impact and bedrock differences on the vegetational history of the Insubrian Southern Alps. Vegetation History and Archaeobotany 9: 175–178.

28.

Grime

(1974) Vegetation classification by reference to strategies. Nature 250 (5461): 26–31.

29.

Hurlbert

(1971) The nonconcept of species diversity: A critique and alternative parameters. Ecology 52: 577–586.

30.

Isbell

Calcagno

Hector

. (2011) High plant diversity is needed to maintain ecosystem services. Nature 477: 199–202.

31.

Kaltenrieder

Procacci

Vanniere

. (2010) Vegetation and fire history of the Euganean Hills (Colli Euganei) as recorded by Lateglacial and Holocene sedimentary series from Lago della Costa (northeastern Italy). The Holocene 20: 679–695.

32.

Lambeck

Antonioli

Purcell

. (2004) Sea-level change along the Italian coast for the past 10,000 yr. Quaternary Science Reviews 23: 1567–1598.

33.

Lang

(1994) Quartaere Vegetationsgeschichte Europas: Methoden und Ergebnisse. Jena: G. Fischer.

34.

Lawson

Frogley

Bryant

. (2004) The Lateglacial and Holocene environmental history of the Ioannina basin, north-west Greece. Quaternary Science Reviews 23: 1599–1625.

35.

Legendre

(2001) Model II Regression – User’s Guide. Montréal: Département de Sciences Biologiques, Université de Montréal.

36.

Legendre

(1988) Numerical Ecology (Developments in Environmental Modelling). 2nd Edition. Amsterdam: Elsevier.

37.

Leighton

(1999) Sicily before History: An Archaeological Survey from the Palaeolithic to the Iron Age. Ithaca, NY: Cornell University Press, 312 pp.

38.

Lotter

(1999) Late-glacial and Holocene vegetation history and dynamics as shown by pollen and plant macrofossil analyses in annually laminated sediments from Soppensee, central Switzerland. Vegetation History and Archaeobotany 8: 165–184.

39.

Lotter

Eicher

Siegenthaler

. (1992) Late-glacial climatic oscillations as recorded in Swiss lake-sediments. Journal of Quaternary Science 7: 187–204.

40.

Lotter

Heiri

Brooks

. (2012) Rapid summer temperature changes during Termination 1a: High-resolution multi-proxy climate reconstructions from Gerzensee (Switzerland). Quaternary Science Reviews 36: 103–113.

41.

Magny

Peyron

Sadori

. (2012) Contrasting patterns of precipitation seasonality during the Holocene in the south- and north-central Mediterranean. Journal of Quaternary Science 27: 290–296.

42.

Malone

(2003) The Italian Neolithic: A synthesis of research. Journal of World Prehistory 17: 235–312.

43.

Marchetto

Colombaroli

Tinner

(2008) Diatom response to mid-Holocene climate change in Lago di Massaciuccoli (Tuscany, Italy). Journal of Paleolimnology 40: 235–245.

44.

Marquard

Weigelt

Temperton

. (2009) Plant species richness and functional composition drive overyielding in a six-year grassland experiment. Ecology 90: 3290–3302.

45.

Medail

Quezel

(1999) Biodiversity hotspots in the Mediterranean basin: Setting global conservation priorities. Conservation Biology 13: 1510–1513.

46.

Millennium Ecosystem Assessment (2005) Ecosystems and Human Well-Being: General Synthesis. Washington, DC: Island Press.

47.

Mooney

Mace

(2009) Biodiversity policy challenges. Science 325: 1474.

48.

Moreno

Oechel

(1991) Fire intensity effects on germination of shrubs and herbs in southern California chaparral. Ecology 72: 1993–2004.

49.

Myers

Mittermeier

. (2000) Biodiversity hotspots for conservation priorities. Nature 403: 853–858.

50.

Odgaard

(1999) Fossil pollen as a record of past biodiversity. Journal of Biogeography 26: 7–17.

51.

Odgaard

(2006) Biodiversity of past plant cover. In: Elias

(ed.) Encyclopedia of Quaternary Science. Amsterdam: Pergamon Press, pp. 2511–2515.

52.

Overpeck

Whitlock

Huntley

(2003) Terrestrial biosphere dynamics in the climate system: Past and future. In: Alverson

Bradley

Pedersen

(eds) Paleoclimate, Global Change and the Future. Berlin: Springer, pp. 81–103.

53.

Pausas

Carbo

Caturla

. (1999) Post-fire regeneration patterns in the eastern Iberian Peninsula. Acta Oecologica: International Journal of Ecology 20: 499–508.

54.

Pereira

Leadley

Proenca

. (2010) Scenarios for global biodiversity in the 21st century. Science 330: 1496–1501.

55.

Pimm

Russell

Gittleman

. (1995) The future of biodiversity. Science 269: 347–350.

56.

Rasmussen

Andersen

Svensson

. (2006) A new Greenland ice core chronology for the last glacial termination. Journal of Geophysical Research: Atmospheres 111(D6). DOI: 10.1029/2005JD006079.

57.

R Development Core Team (2008) R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing. Available at: http://www.R-project.org (accessed 21 November 2012).

58.

Reich

Tilman

Isbell

. (2012) Impacts of biodiversity loss escalate through time as redundancy fades. Science 336: 589–592.

59.

Samartin

Heiri

Vescovi

. (2012) Lateglacial and early Holocene summer temperatures in the southern Swiss Alps reconstructed using fossil chironomids. Journal of Quaternary Science 27: 279–289.

60.

Schwander

Eicher

Ammann

(2000) Oxygen isotopes of lake marl at Gerzensee and Leysin (Switzerland), covering the Younger Dryas and two minor oscillations, and their correlation to the GRIP ice core. Palaeogeography, Palaeoclimatology, Palaeoecology 159: 203–214.

61.

Seppä

(1998) Postglacial trends in palynological richness in the northern Fennoscandian tree-line area and their ecological interpretation. The Holocene 8: 43–53.

62.

Soepboer

Sugita

Lotter

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

63.

Stoate

Baldi

Beja

. (2009) Ecological impacts of early 21st century agricultural change in Europe – A review. Journal of Environmental Management 91: 22–46.

64.

Sugita

(2007) Theory of quantitative reconstruction of vegetation I: Pollen from large sites REVEALS regional vegetation composition. The Holocene 17: 229–241.

65.

Sugita

Parshall

Calcote

(2006) Detecting differences in vegetation among paired sites using pollen records. The Holocene 16: 1123–1135.

66.

Terral

Alonso

Capdevila

RBI

. (2004) Historical biogeography of olive domestication (Olea europaea L.) as revealed by geometrical morphometry applied to biological and archaeological material. Journal of Biogeography 31: 63–77.

67.

Thuiller

Lavorel

Araujo

. (2005) Climate change threats to plant diversity in Europe. Proceedings of the National Academy of Sciences of the United States of America 102: 8245–8250.

68.

Tinner

Conedera

Ammann

. (1998) Pollen and charcoal in lake sediments compared with historically documented forest fires in southern Switzerland since AD 1920. The Holocene 8: 31–42.

69.

Tinner

Conedera

Ammann

. (2005) Fire ecology north and south of the Alps since the last ice age. The Holocene 15: 1214–1226.

70.

Tinner

Conedera

Gobet

. (2000) A palaeoecological attempt to classify fire sensitivity of trees in the southern Alps. The Holocene 10: 565–574.

71.

Tinner

Hubschmid

Wehrli

. (1999) Long-term forest fire ecology and dynamics in southern Switzerland. Journal of Ecology 87: 273–289.

72.

Tinner

Lotter

Ammann

. (2003) Climatic change and contemporaneous land-use phases north and south of the Alps 2300 BC to 800 AD. Quaternary Science Reviews 22: 1447–1460.

73.

Tinner

van Leeuwen

JFN

Colombaroli

. (2009) Holocene environmental and climatic changes at Gorgo Basso, a coastal lake in southern Sicily, Italy. Quaternary Science Reviews 28: 1498–1510.

74.

Turner

Roberts

Jones

(2008) Climatic pacing of Mediterranean fire histories from lake sedimentary microcharcoal. Global and Planetary Change 63: 317–324.

75.

van der Knaap

(2009) Estimating pollen diversity from pollen accumulation rates: A method to assess taxonomic richness in the landscape. The Holocene 19: 159–163.

76.

van der Knaap

Ammann

(1997) Depth-age relationships of 25 well-dated Swiss Holocene pollen sequences archived in the Alpine Palynological Data-Base. Revue de Paléobiologie 16: 433–480.

77.

Vanniere

Colombaroli

Chapron

. (2008) Climate versus human-driven fire regimes in Mediterranean landscapes: The Holocene record of Lago dell’Accesa (Tuscany, Italy). Quaternary Science Reviews 27: 1181–1196.

78.

Vescovi

Ravazzi

Arpenti

. (2007) Interactions between climate and vegetation during the Lateglacial period as recorded by lake and mire sediment archives in Northern Italy and Southern Switzerland. Quaternary Science Reviews 26: 1650–1669.

79.

Wagner

Lotter

Nowaczyk

. (2009) A 40,000-year record of environmental change from ancient Lake Ohrid (Albania and Macedonia). Journal of Paleolimnology 41: 407–430.

80.

Wehrli

Tinner

Ammann

(2007) 16 000 years of vegetation and settlement history from Egelsee (Menzingen, central Switzerland). The Holocene 17: 747–761.

81.

Weng

Hooghiemstra

Duivenvoorden

(2006) Challenges in estimating past plant diversity from fossil pollen data: Statistical assessment, problems, and possible solutions. Diversity and Distributions 12: 310–318.

82.

Willis

Birks

HJB

(2006) What is natural? The need for a long-term perspective in biodiversity conservation. Science 314: 1261–1265.

83.

Winter

Schweiger

Klotz

. (2009) Plant extinctions and introductions lead to phylogenetic and taxonomic homogenization of the European flora. Proceedings of the National Academy of Sciences of the United States of America 106: 21,721–21,725.

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

0.20 MB