Middle- to late-Holocene fire history and the impact on Mediterranean pine and oak forests according to the core RF93-30,central Adriatic Sea

Abstract

The high-resolution Adriatic RF93-30 core shows changes in its microcharcoal record, which correlate to terrestrial fires from the last 7000 years. Pollen and microcharcoals were transported by wind and fluvial transport from the sedimentary basin, including the Po River and other rivers flowing into the sea off the Italian east coast. Charcoal particles and pollen were counted in the same samples, and the maximum breadth and length of charcoal particles were measured. Microcharcoals with large dimensions were taken as fire indicators occurring along the near coast, as they probably arrived from short distances, the nearest being in Apulia, in southern Italy. The age–depth model was developed within the multidisciplinary PALICLAS project. Several potential fire activity increases (PFAIs) were visible as peaks in the diagram. The oldest PFAIs occurred at the middle Holocene (approximately dated to c. 6730, 5430, 4150 cal BP), others occurred at the late Holocene (c. 3760, 2660, 2240, 2030, 1930, 1510 cal BP) and during the last millennium (c. 900–865, 530, 120–96 cal BP). The two oldest peaks in the diagram, occurring in the 7th–6th millennia, showed the highest contribution of charcoal corresponding to the highest values of arboreal pollen (AP) in the sedimentary record. Although the CHAR peaks did not represent a single fire event, the diagram suggests a good correspondence between paleofire activity and terrestrial vegetation biomass during this early phase. Pollen containing black particles was observed, which suggested some grains were transported in suspension with winds from burned woods. The main unambiguous anthropogenic fire causation would have occurred during the last four millennia. From 4.2 ka, it became hard to disentangle climate and Bronze Age actions. Technology and human activity probably improved the pace of fire events, especially involving oak woods, with evidence of an increase of CHAR during the last millennium.

Keywords

archaeology Bronze Age climate dark-filled pollen human impact pollen

Introduction

Regardless the trigger, fire significantly affects vegetation cover, biodiversity, and soil composition, and it produces socioenvironmental consequences that can be observed at decadal to centennial scales in paleoenvironmental records (Vannière et al., 2008). In the middle Holocene, global warming and human impact have potentially played synergistic roles in eliciting the frequency and amplitude of fires in Mediterranean regions, which are characterized by mosaic ecosystems and high population density (Burjachs and Expósito, 2015; Sadori et al., 2013a). Vegetation history demonstrates that the plant biomass has continuously varied while archaeological surveys show that the demography experienced oscillations, but had substantially increased from prehistory to history (Palmisano et al., 2017; Piovesan et al., 2018). The study of fire history includes knowledge on the diachronical changes in vegetation types, land-cover flammability and conductivity, elevation, and slopes, which are all variables that are useful in modelling the spatial-heterogeneous Mediterranean landscape together with climate parameters (seasonality, wind, and average moisture; Conedera et al., 2018; Millington et al., 2009). The cooperation of ecology with paleoecology, thereby combining knowledge on past and current phenomena, aids in the consideration of phenomena on a long-term perspective (Marignani et al., 2017; Mercuri et al., 2015b; Rull, 2010). Therefore, understanding the causes and pace of fire occurrence in past environments may be useful in assessing the impact of human actions on the status of current vegetation.

Most of the research on past fire dynamics is based on the study of charcoal of different sizes and dimensions. According to Marquer (2010), who studied archaeological contexts, charcoals can be distinguished as being macrocharcoals (>500 μm), mesocharcoals (500–160 μm), or microcharcoals (<160 μm). The macrocharcoals are pieces of charcoals in soils (pedoanthracology: Nelle et al., 2013; charcoal kilns: Benatti et al., 2018) and in stratigraphic series of both sediments (sediment anthracology; Carracedo et al., 2018) and archaeological layers (archaeobotany; Fiorentino and Magri, 2008; Mercuri et al., 2019). Meso- and microcharcoals are small/very small fragments in sediments and archaeological layers (Cui et al., 2009; Damnati and Reddad, 2017; Peresani et al., 2018; Sadori, 2018). Microscopical charcoal fragments in sedimentary records are excellent markers of fire occurrence in a given area and over a long-term period because the vegetation burned during extended fires produces small carbonized particles that are resistant to decomposition and may be transported from the combustion site to the sedimentation basin (Jones et al., 1997; Verardo, 1997). Paleoenvironmental reconstructions based on high-resolution analyses of sediments, including microcharcoals from terrestrial cores, are quite common in the Mediterranean basin (central: Colombaroli et al., 2008; Kaltenrieder et al., 2010; western: Carracedo et al., 2018; Reddad et al., 2012; eastern: Turner et al., 2008).

Besides the terrestrial records, sedimentary marine cores are exceptional records for regional fire history as they are far from the local signal of archaeological sites. Depending on the wind and/or river flow, the transportation and sedimentation of microcharcoals are similar to that of pollen, deposited over a short range of time, at most a few years, on the ocean floors. Moreover, it is assumed to not have a significant time lag between production and deposition (see references for marine environments in Daniau et al., 2007, 2010). The microcharcoal analysis of sedimentary sequences from marine cores MD95-2042 and MD04-2845, dated between 70 and 10 ka, were connected to the natural fire history in the areas of the terrestrial hydrographic basins in western Europe (SW Iberia). Biostratigraphical data suggest that the fires were mainly driven by the millennial-scale climatic variability of Dansgaard–Oeschger events consisting of rapid warming episodes followed by gradual cooling. Variability of temperatures and precipitation has also had an impact on vegetation, and the consequent biomass availability resulted in the variability of natural fires and by-products emission in the environment (Daniau et al., 2007, 2010). Marine sediments from off-shore of Cape Blanc, West Africa, have helped to delineate the role of fires at the dry limit of the savannah during Holocene oscillations showing that fires increased during the extension of plant cover and wet phases of the early Holocene, while at the end of the middle Holocene, maximum fire occurrences accelerated the desertification of the southern Sahara (Dupont and Schefuß, 2018).

In the Mediterranean basin, at the transitional marine-land zones in Sardinia, vegetation and fire dynamics were related to the establishment of the Mediterranean climate variability. Climatic features included moist and cool summers with dry and mild winters specifically occurring in the NE coast after 5300 cal BP (Stagno di Sa Curcurica; Beffa et al., 2016), and it triggered an increase of fire frequency from 4650 cal BP in the central west coast (Mistras Lagoon; Di Rita and Melis, 2013).

Located in the Adriatic Sea, a few kilometres from the southern Italy eastern coast, the core RF93-30 represents an excellent off-site record that is useful to investigate fire history on a wide regional scale, including eastern North to South continental Italy. A previous palynological study (Mercuri et al., 2012) highlighted the unique potential of this marine core for investigating the development of cultural landscapes under changing climate and human impact during the middle and late Holocene in the central Mediterranean. Interdisciplinary research has shown a significant correlation between the nonarboreal pollen percentages and a mineral magnetic parameter that was interpreted as being representative of a high input of sediments from erosional processes as a human impact signal, which was clearly visible for the first time in marine cores (Oldfield et al., 2003). Selected pollen taxa were compared with lake and archaeological records to trace the development of cultural landscapes characterized by cultivated trees (OJC in Mercuri et al., 2013b) and the most frequent anthropogenic pollen indicators (API in Mercuri et al., 2013a). The trend of trees over the last 2300 years was studied for conservation purposes revealing an inverse correlation between tree biomass and population density in the Italian peninsula (Piovesan et al., 2018). In this article, we reconsider the vegetation dynamics, including the microcharcoal fire signal in sediments to search for a correlation between tree decrease/forest dynamics and fires over the last 7000 years. New evidence on charcoal peaks, interpreted as signals of potential fire activity increase (PFAI), from the eastern side of the Italian peninsula is provided. According to Whitlock and Larson (2001), the interpretation of charcoal peaks in sediments must take into consideration the following assumptions: (i) When stratigraphic levels incorporate abundant charcoal, the palynological diagrams show ‘charcoal peaks’ that are inferred to be evidence of past fires. (ii) A charcoal peak is composed of particles deposited both during and after a fire, because the deposition of charcoal particles can occur from a few years to decades after the actual fire. (iii) This means that the presence of a charcoal peak alone is not sufficient to infer the levels of fire intensity or fire size in the past based on charcoal abundance in the stratigraphic record. (iv) Nevertheless, large charcoal particles can be associated with high intensity fires where turbulent winds move such particles beyond the combustion zone.

In this article, we describe the charcoal record in a sedimentary core with special consideration of the charcoal peaks observed in the palynological diagram. They are not unambiguous evidence of one specific event, due to the copresence of primary and secondary charcoal describing several years in the same layer, but rather they give good evidence of the occurrence of fires during the range of years described by each sample. The use of ‘charcoal peaks’ is simple but not simplistic and allows a comparison of similar evidence at similar chronologies found in other terrestrial records of the Mediterranean basin.

The RF93-30 marine core includes biological particles arriving from a wide continental area, and therefore, the presence of both climate and human-driven events is expected. An attempt to disentangle the different forces eliciting fires is made although it is well known that, despite a dry climate with lower biomass has induced a decrease of fire-carbon emissions since 7 ka BP, the human presence has been able to double the fire frequency in the South European continent over the whole Holocene (Vannière et al., 2016).

Methods

Core RF93-30, located 18 km north of the northern coast of the Gargano promontory and 55 km South-East of the center of the Mid-Adriatic Depression, was collected from a shallow shelf at a water depth of 77 m by the CNR–IGM teams from Bologna (Trincardi et al., 1996; Figure 1). The core is approximately 627 cm long and was analyzed with a multiproxy approach as part of the PALICLAS project (Lowe et al., 2007; Oldfield et al., 2003). Stratigraphic data indicate that there were strong links between land cover, surface processes, and near-shore marine sedimentation suggesting that the main sediment supply basin was the Po Valley. Globigerina sacculifer, an oligotrophic planktonic foraminifera typical of warm, tropical environments, and that lives today in the West Mediterranean Sea at the end of summer, provided key biostratigraphical information on climate: its main peaks during the last eight millennia of sea-level high stand are indicative of relative climatic optimum with low turbidity of the water column and reduced river runoff; reversely, its minima are interpretable as index of cool and rainy/wet events (Piva et al., 2008). In addition, the opportunistic benthic Valvulineria complanata is an interesting foraminifer, which lives in the mud belt environment in shallow shelf settings, characterized by high levels of organic matter and river runoff; as marker of cold and humid intervals, this species was taken as representative of environmental conditions during the ‘Little Ice Age’ (LIA; Piva et al., 2008).

Figure 1.

Location map of core RF93-30, and Italian sites quoted in the article: 1. Core RF93-30; 2. Lago Battaglia; 3. Lago Alimini Piccolo; 4. Lago Salso; 5. Grotta delle Mura; 6. Lago di Monticchio; 7. Terramara di Montale; 8. Terramara di Baggiovara; 9. Terramara di Poviglio; 10. Torbiera di Pavullo; 11. Lago di Massaciuccoli; 12. Lago dell’Accesa; 13. Lago Lungo, Rieti basin; 14. Lago di Mezzano; 15. Valle di Castiglione; 16. Mistras Lagoon; 17. Lago di Pergusa; 18. Biviere di Gela.

Pollen samples were collected at c. 10 cm intervals, from 4 cm to 622 cm depth (63 samples). The estimated sedimentation rate was comparatively lower in the bottom part (from 622 to 527 cm: 0.29 mm yr^–1) and higher in the other parts (from 527 to 365 cm: 1.72 mm yr^–1; 365–219 cm: 0.97 mm yr^–1; 219–129 cm: 1.61 mm yr^–1; 129–0 cm: 1.98 mm yr^–1; Langone et al., 1996).

The age model shows that the core approximately covers the last 7000 years and was mainly based on a combination of stratigraphical and biostratigraphical correlations with other marine cores (Lowe et al., 1996, 2007; Trincardi et al., 1996), paleomagnetism, the Avellino tephra (AT) at 529 cm and two ¹⁴C dates on foraminifers. Recently, the age-depth model (Figure 11 in Oldfield et al., 2003) was revised and a new age-depth model, slightly different from the previous one, was developed based on the magnetic inclination record (Figure 2 by L. Vigliotti in Mercuri et al., 2012). The temporal resolution of each sample was calculated to ~300–100 years in the bottom part (until 365 cm depth per 2240 cal. BP) and then decreases to ~40–20 years per sample, given the high sedimentation rates at the core site, which makes possible to distinguish centennial to decadal curve oscillations.

Figure 2.

Charcoal accumulation rate and total charcoal concentration from the samples of core RF93-30.

Pollen and charcoal were extracted using a standard pollen extraction procedure including sieving with a 7 µm nylon mesh and floatation with sodium metatungstate hydrate (specific gravity: 2.0 g cm^–3; see description in Florenzano et al., 2012). The chemical processing could imply some fragmentation of charcoals, but this should not result in an underrepresentation of local fires (Sadori and Giardini, 2007) as it is obvious that we cannot have fires in the close vicinity of the sampling point in the sea. Lycopodium tablets were added to estimate CHAC, the charcoal particles concentration expressed as charcoal per unit of volume (number of charcoals cm^–3). The CHAC values were multiplied by the estimated sedimentation rate (cm years^–1) to calculate the charcoal accumulation rate CHAR (number of charcoals·cm^–2 years^–1). Pollen data are reported as percentage and concentration (pollen cm^–3).

Charcoal particles have been identified and counted in the same samples of the pollen record (according to Sadori and Giardini, 2007; Tinner and Hu, 2003). Two charcoal-size parameters were measured: the maximum breadth and the maximum length, and the obtained measures were sorted into three size classes (10–50 µm; 50–125 µm; > 125 µm). We call ‘large microcharcoals’ the items > 125 µm to avoid ambiguous terminology. Particles < 10 µm were not considered as they can easily result from fragmentation, while the other size classes may be evidence of regional fires and useful to roughly estimate the distance of the fire (Whitlock, 2001).

Pollen percentages were calculated on the total pollen counts. The selected pollen groups are OJC (Olea + Juglans + Castanea; Mercuri et al., 2013b), API (Artemisia + Centaurea + Plantago + Urtica + Trifolium type; cereals and Cichorieae are plotted separately due to their respective specific value of crop and pasture indicators; Florenzano et al., 2015; Mercuri et al., 2013a), thermo-Mediterranean taxa (Juniperus type, Quercus ilex type, Rhamnus type, Pistacia, Myrtus; Mariotti Lippi et al., 2018). Olea europaea is excluded from the latter sum for its double role of wild and cultivated tree (Languut et al., 2019). Following Sadori and Giardini (2007), the ratio CH/AP concentrations (CHarcoal/Arboreal pollen, Figure 4) is calculated as a further support to evaluate past fires dynamics. Tree pollen taxa are plotted to show parallel or opposite trends of the main forest biomass acting as source (fuel) for burning.

Results

Size classes

As expected, most of microcharcoals belong to the class 10–50 µm both considering the breadth (99.0% of the items) and the length (87.3%; Figure 2). Other items belong to the class 50–125 µm for length (11.4% of the items), and few for breadth (1.0%). A few large microcharcoals, with length > 125 µm, were observed (1.3%).

Total charcoal concentration

Microcharcoals were observed in all samples of the marine core showing c. 1,20,000 ch·cm^–3 on average, with the variable total charcoal concentration (T CHAC), from c. 21,150 ch·cm^–3 (at 322 cm, dated to approximately 1823 cal BP) to c. 918,000 ch·cm^–3 (at 612 cm, 6723 cal BP).

The CHAR values, which are the same values corrected from changes in the sedimentation rate, reached 23,000 ch·cm^–2 yr^–1 in the same sample. CHAR is c. 17,000 ch·cm^–2 yr^–1 on average, and ranged from c. 1240 to c. 1,59,400 ch·cm^–2 yr^–1.

The second highest charcoal concentration was again found in a middle-Holocene sample of the bottom part of the core (572 cm, 5427 cal BP), was calculated as 6,61,500 ch·cm^–3, resulting in a CHAR of 16,500 ch·cm^–2yr^–1. Then, other peaks of the T CHAC and CHAR were visible in the diagram (Figure 2). Despite the different values, the CHAC and CHAR curves showed similar trends, with CHAR values lowered by the low sedimentation rate in the bottom part (Figure 3).

Figure 3.

Microcharcoal diagram of core RF93-30.

Trends of AP

The percentage of trees and shrubs was significantly high (76–88%) in the bottom half of the diagram (Figure 4). A decreasing trend began just after 3760 cal BP (at 512 cm) when the AP sum falls from 78% to 70% (at 502 cm, 3620 cal BP). After this point, the curve had lower values with oscillations, and a minimum of 53% was reached at 2763 cal BP (422 cm). The AP sum reached its last maximum of 85% at 1510 cal BP (312 cm). Then, its trend continued to decrease in a rather irreversible way touching minimum values < 50% in the last century, toward the top of the diagram.

Figure 4.

Percentage pollen diagram of core RF93-30: selected trees and shrubs (exaggeration × 5).

Pollen grains filled with black particles

Interestingly, in the bottom samples between 622 and 542 cm, from c. 7050 to c. 4460 cal BP, a number of very well-preserved pollen grains filled with black particles, with different features and probably different origin, were observed. No brown fragments suggesting organic matter were observed in the grains. Some black inclusions resemble the charcoal particles recovered in the same slides. Others could be pyrite, which can form under reducing conditions during the bacterial breakdown of organic material in the sediments, but the heavy liquid we used during pollen extraction should have definitively removed the heavy pyrite particles from our samples (Proske et al., 2015). Pollen with black particles belonged to oaks (Quercus ilex type and deciduous Quercus) and also Pinus, Fagus, and rarely, shrubs and herbs (Amaranthaceae-Chenopodiaceae; Figure 5).

Figure 5.

Pollen filled with black particles: 1. Pinus (80 µm, slide no.116); 2. Pinus (76 µm, slide no.117); 3. Fagus (35 µm, slide no.117); 4. Quercus ilex type (25 µm, slide no.125); 5. Quercus ilex type (29 µm, slide no.117); 6. Quercus ilex type (28 µm, slide no.117); 7. Quercus ilex type (28 µm, slide no.117); 8. Quercus ilex type (29.5 µm, slide no.125); 9. and 10. Chenopodiaceae (22 µm; slide no.125); 11. Ostrya carpinifolia/Carpinus orientalis (24 µm, slide no.114). The chronology of samples with these pollen grains is c. 7070 cal BP (slide no.125), c. 5750 cal BP (slide no.117), c. 5430 cal BP (slide no.116), c. 5100 cal BP (slide no.114).

Discussion

Palynological analysis of the core RF93-30 showed heterogeneous vegetation communities from northern to southern Italy, highlighting that pollen inputs emanated both from river flows and airborne transport (Mercuri et al., 2012). Similarly, the microcharcoals in the sediments of this marine core would have been transported partly by wind from the vegetation, especially grown in front of the coring site, and partly by fluvial transport by the Po River and other rivers flowing into the sea off the east coast. A good index of input of alluvial deposits was considered the presence of reworked pollen in sediments from erosional processes; this presence was found to be always important (> 40% of reworked pollen in all samples), but significantly lower in the bottom part of the core (Figure 3 in Mercuri et al., 2012).

The dominance of charcoal items belonging to the smallest class (10–50 µm) was expected due to the distance of the core from the coast (at least 18 km) and the obvious absence of plant cover in situ. The items with larger dimensions seemed to have arrived from short distances, the nearest being the coast of Apulia, southern Italy. In fact, the 125–160 µm microcharcoals were also readily transported by wind, even at a low velocity (Clark, 1988). When observed, the synchronous trends of curves of the different size classes of microcharcoals could be interpreted as evidence of similar fire dynamics in the regional and local/nearest occurrences (Caroli and Caldara, 2007).

In core RF93-30, charcoal data suggested the existence of fires over different lands even if the prevalence of wind or flooding transport of charcoal particles to the core sediments could not be estimated. The very good preservation of pollen in the oldest samples, including those containing black particles in samples rich of microcharcoals (Figure 5), suggested that some dust arrived by winds, most easily transported in suspension with (South-)West-East circulation. We suggest that part of the pollen grains may have been filled with smoke, and that, if this happened, the flames did not touch the pollen but the grains would have been airborne due to intense heat waves and wind. This interpretation suggests that there was a simultaneous transport of pollen and ash (like a black haze?) from the site where the fire occurred to the sea. The smoke and heat generated by dramatic fire events might have lifted the pollen grains from burning plants in the warm and black haze, emptying them out, and then filling them with the smallest black particles arriving from the coast. Today, hot air from exceptionally hot fires can result in whirls, creating powerful tornados of flames and wind (e.g. this was the case of the dramatic ‘firenado’ in California in 2018).

The peculiar context of the marine core encouraged us to point our attention to the charcoal peaks, observed as T CHAC and CHAR. Following the sedimentation rate, the peaks in charcoal concentration were more evident in the oldest samples while the accumulation rate peaks increased toward the last part of the record (Figure 2). This was expected, as the earliest CHAR peaks corresponded to a time of sedimentation that was estimated at about 200 years in the oldest samples, which was approximately 10 times higher than the time estimated for the more recent samples.

Therefore, the main peaks visible in the pollen diagram are discussed below as PFAI to outline that they may have been a result of several inputs from subsequent years and adjacent terrestrial regions, and that a peak may have represented one or more fires occurring during the time span represented by the sample. Moreover, the presence of ‘noise effects’ from analysis procedures or random variations in sediments could not be definitively excluded. All these issues were considered to infer a paleoenvironmental interpretation that results from the evidence of microscopical charcoal and pollen from this core, which combines such evidence with known data from terrestrial cores and archaeological sites of the Italian peninsula.

The chronology of the different PFAI in the marine core

Below, a number of PFAIs from the core RF93-30, which also outline increases of large microcharcoals, are described and discussed in chronological succession. We report the biostratigraphical information and main inferences on the possibility of climate or humans responsibility in triggering fires at the different times, with rough discrimination based on three large chronological ranges from the oldest middle Holocene to the early late Holocene, and, then, the fire events over the last millennium in the marine record.

The middle Holocene PFAI (from c. 6730 to c. 4150 cal BP)

c. 6730 cal BP (4780 BC) – Neolithic extended fires in SE Italy

This is the highest value for the total concentration of charcoal particles (T CHAC), also recorded as the increase in the CHAR record, including the highest value of large microcharcoals (> 125 µm) in the diagram. However, the total pollen concentration was relatively low; the percentage of trees had a slight decrease matching an increase in shrubs, especially Juniperus type (Figures 4 and 6); a percentage decrease was also visible in Pinus, deciduous Quercus, and other tree curves (Acer, Fraxinus, Tilia), while Carpinus orientalis/Ostrya carpinifolia, Carpinus betulus, and Corylus increased in the mesophilous oak woods. Thin increases were visible in Alnus and Salix near wetlands, and Betula, Taxus, and Fagus in the mountain belt. Quercus ilex type and Mediterranean shrubland (Juniperus type and Phillyrea) increased. Cereals and Cichorieae had a slight increase. Meanwhile, the pollen grains filled with charcoal particles observed at this level suggested large and violent fires.

Figure 6.

Concentration and percentage pollen diagram of core RF93-30: CH/AP ratio, selected pollen sums, and taxa.

Inferences

Fires occurred in a centennial phase without a very high biomass (forest and shrubland cover were low and even decreased) suggesting that there was a limited fuel supply. Fires probably occurred in the near coast (allowing the arrival of large microcharcoals by wind) and involved mesophilous oak woods giving space to other types of woods or open lands. The microcharcoal record had a coeval high peak, including large microcharcoals dated to c. 6800–6400 cal BP at Lago Battaglia (core BAT1/2 in Figure 6 by Fiorentino et al., 2013 quoting Caroli, 2005), in the Gargano coast facing our Adriatic core. This strongly supported the evidence that the Gargano region was affected by (repeated?) fires during the early middle Holocene.

In terrestrial records, there was the presence of several archaeological sites along the near coast and surrounding region (see Tables 1 and 2 listing Neolithic sites, from 8400 cal BP in Fiorentino et al., 2013), and signs of human activities that could archaeologically be dated to the recent Neolithic (c. 6800–6300 cal BP; Fiorentino et al., 2013). The spot presence of cereal pollen in the RF93-30 core at this phase suggests farmers might have been involved in setting fires to mesophilous oak woods for field cultivation, and even daily life activities. The action should have given space to the spreading of Mediterranean oak and shrub lands. Although the OJC curve was absent from the marine core, signs of the fairly continuous curve of Olea began soon after, at c. 6000 cal BP, in the terrestrial cores of Apulia (Lago Salso and Lago Alimini Piccolo: Di Rita et al., 2011; Di Rita and Magri, 2009; Lago Battaglia: Caroli and Caldara, 2007).

c. 5430 cal BP (3480 BC) – Regional or far (climatically driven) fires

This was the second peak in the amplitude of T CHAC, which was also recorded in the CHAR record. However, large microcharcoals were absent, and total pollen concentration had the highest value, while the AP percentage was still very high. Among the trees, several mesophilous and wetland trees decreased (Ulmus, Acer, Humulus, and Salix). However, Pinus had its last major peak, which was significant, as well as that of the deciduous Quercus. Then, pines and oaks started to decline, while Quercus ilex type continued to expand after this point. Cereals were present in traces, and Cichorieae were steady.

Inferences

Fires occurred during a phase of high biomass value (forest cover) and involved different types of mixed woods, including hygrophilous woods. Although the fires were not spread in the near coast, some human activity was evident from pollen. A minor peak of microcharcoals without large particles was also recorded at Lago Battaglia (Caroli and Caldara, 2007) suggesting the occurrence of fires at a regional level. The presence of late recent Neolithic, and then Chalcolithic/Copper Age populations, that continued to exploit woods for fortified villages apart from agriculture and daily life activities, is well known in Italy (e.g. Arene Candide, Nisbet, 2008; Rottoli and Castiglioni, 2009; Defensola A and Scaloria in Gargano, Fiorentino et al., 2013). Nevertheless, most signals of fire during this period seem to not have been truly ‘specialized’ or ‘culturally-oriented’ as pines and oaks (that are trees largely used for building and other uses in the past; see, for example, Mercuri et al., 2019) show high values in the pollen diagram. Plants seem to have been burnt in a rather random way with no clear evidence of plant selection.

The role of humans cannot be excluded, as the fire episode(s) probably followed a quite general high biomass and forest cover abundance dynamics within natural events. Climate change and some vulcanological events that occurred at these times would have influenced the distribution and richness of forest cover in large areas of the terrestrial environment (see below).

c. 4150 cal BP (2200 BC) – The drought peak that is thought to have triggered cultural changes

There was an increase of T CHAC and CHAR, though it was less evident than the two previous fire peaks and was without large microcharcoals, and a high total pollen concentration. This occurred after a 400-year-long phase of fires along the coast, which was revealed by the significant presence of large microcharcoals between c. 4800 and c. 4400 cal BP, and signs of local fires in the Lago Battaglia record (core BAT4 in Caroli and Caldara, 2007; Figure 7). The RF93-30 core showed that an irreversible decline of pines started during this period, and gave space to the increase of Quercus ilex type and a gradual increase of Cichorieae. Then, during the fire peak at 4150 cal BP, the tree percentage was still high; deciduous Quercus and other mesophilous trees (Carpinus orientalis/Ostrya carpinifolia, Ulmus) increased while Abies and Pinus reached a minimum. Moreover, Quercus ilex type had a significant decrease together with Juniperus type, while Phillyrea and Pistacia cf. lentiscus were present among the Mediterranean plants. Cereals are low and Cichorieae decreases.

Figure 7.

Selected data from core RF93-30 compared with AP percentage sum and microcharcoal concentrations of Lago Battaglia (core BAT1; Caroli and Caldara, 2007) and Lago Alimini Piccolo (Di Rita and Magri, 2009).

Inference

This occurred with the 4.2 ka, which was a highly significant date for climate (Bond et al., 1997) and archaeological (Weiss et al., 1993) reasons and was characterized by complex intermingled processes (Bini et al., 2019; Di Rita and Magri, 2019). Between 5000 and 4000 cal BP, with latitudinal shifts, the climate became increasingly arid, also causing a crisis of some civilizations in the eastern Mediterranean (Mercuri et al., 2011), and culminating in the drought event 4.2, a W-shaped c. 300 years long event, especially visible in the drop in rainfall in the Levant (Kaniewski et al., 2018). Despite this, in the Adriatic cores, including the RF93-30, minimum of G. sacculifer marks a cool (and rainy?) event at 4100–3800 cal BP (Piva et al., 2008). The PFAI recorded in RF93-30 preceded the cool event and occurred in a phase of relatively high biomass value (although forest cover had an overall decrease and lowering of plant biomass during the middle Holocene; Vannière et al., 2016). Fires occurred regionally, and although large microcharcoals were absent from core RF93-30, there were signs of local fires in the Lago Battaglia record (Figure 7). We suggest that these fires were not so violent as the older ones, or the winds did not transport the charcoal particles because of a different circulation or a different season of fires. The decrease of holly oak in our record suggests that evergreen oak woods were largely burnt giving space to Mediterranean shrubland, while signs of human activities were not obvious at this time. Similarly, at Lago Battaglia, the curve of evergreen oaks showed a sharp decline at c. 4200 cal BP, and woodland loss together with a change in fire frequency, not associated with an increase of anthropogenic indicators, became irreversible (Caroli and Caldara, 2007). Moreover, at Lago Alimini Piccolo, a decrease of evergreen vegetation was recorded between 4350 and 3900 cal BP, with a fall of the AP sum corresponding to a rise of microcharcoals (Di Rita and Magri, 2009; Figure 7). In RF93-30, despite the spread of the Early Bronze Age civilizations, the APIs seemed quite low, and conversely, deciduous oaks still had high percentages in the pollen diagram. These trees were emblematic of tree exploitation and continued to increase for about 100 years after this point suggesting a low impact of human activities on the woods. At c. 3900 cal BP, the rise of open landscapes was matched by the very similar curves of Poaceae, Cyperaceae, and Cichorieae (Figure 6), the latter strongly suggesting the expansion of pasturelands under the increasing pressure of pastoralism (Florenzano, 2013; Florenzano et al., 2015).

The early Late Holocene PFAIs (from c. 3760 to c. 1510 cal BP)

c. 3760 cal BP (1210 BC) – The wood exploitation and land management of the Bronze Age people

There was a new increase of T CHAC and CHAR, with a peak of large microcharcoals and total pollen concentration (Figure 3). The tree percentage was high, but showed a rapidly decreasing trend, and the shrub percentage fell. Mountain trees (such as Fagus and Betula) tended to expand with Pinus, which had a peak. Deciduous Quercus and other mesophilous trees (Fraxinus, Corylus) notably decreased, while others increase (Carpinus betulus) or remain steady in the oak woods. Both Quercus ilex type and Juniperus type started to increase; Cichorieae increased, and cereals were absent though their curve became evident and even continuous after this time. From this point, the CH/AP ratio started to increase; this may have simply depended on the decrease of AP or may have suggested a higher amplitude and frequency of the fires.

Inference

Fires occurred in a phase of reduced biomass and forest cover and were also present close to the coast. Fires involved oak woods, while thermo-Mediterranean vegetation was favored to spread; signs of human activities were initially weak. This seemed to have been triggered by a concomitant change of climate and human interference in vegetation dynamics. The 3.7 PFAI opened to a period of increased instability or changes in forest composition under the increasing management of trees and land use exploitation established by several Bronze Age cultures (e.g. in the Po plain; Mercuri et al., 2015c). The pollen diagram showed the start of irreversible trends: oak decrease and fire increase, with quite ‘regular’ oscillations of the CH/AP ratio. The period between c. 3400 and c. 2960 cal BP (1450–1010 BC) seemed to have been particularly characterized by several fire events mirroring sophisticated practices and consolidated knowledge of land management by the Bronze Age cultures (see, for example, the burnt trunk of oak wood at the bottom floor of the Terramare settlement: Figure 2a in Cremaschi et al., 2016; Mercuri et al., 2006b).

Most of the impact on the oak woods seemed to have been made by cutting and wood management by the Middle Bronze Age people. Interestingly, a peak of T CHAR of 1,89,000 ch·cm^–3 (and CHAR of 16,250 ch·cm^–3·yr^–1) occurred at the extreme decrease of trees (AP = 57% with a decrement of > 10% from the 68% of the previous sample, dated to the c. 3400 cal BP). This PFAI was dated to 3285 cal BP and could be regarded as the occurrence of destructive fire events far from the coast, at the time of the Terramare’s crisis in the Po plain (Cremaschi et al., 2016; see also below).

c. 2660 cal BP (710 BC) – The mixed signs of climate and human impact

This peak is visible in the T CHAC and CHAR records, with a significant component of large microcharcoals and total pollen concentration. The tree percentage increased again, while the shrub percentage fell, with steppic Artemisia and Chenopodiaceae falling to zero (Figure 6). The tree increase was due to a temporary rise of deciduous Quercus and some other mesophilous trees (e.g. Fraxinus), while other pollen decreased (Carpinus orientalis/Ostrya carpinifolia) or remained steady in the oak woods. Mountain trees (Picea, Betula, Fagus) decreased, except for Pinus, Abies, and Taxus. Moreover, Alnus decreased in the hygrophilous woods, as well as Quercus ilex type in the thermo-Mediterranean communities. Cichorieae and cereals had a slight decrease as well as the other APIs.

Inference

Fires occurred in a phase of increased biomass and forest cover, and probably also on the coast. They possibly involved some mountain/hilly regions and holly oak woods; human activities were reduced. The OJC curve, still mainly constituted of olive tree pollen, fell to zero suggesting that the role of tree cultivation was not significant and confirming that fires affected the thermo-Mediterranean vegetation.

Before this PFAI, a minimal G. sacculifer (3200–2800 cal BP; Figure 7) during a peak of V. complanata indicated a cool and wet phase (Piva et al., 2008). At that time, Lago Battaglia shores hosted many chenopods that suddenly fell at c. 2700 cal BP, when the maximum microcharcoal value was observed in the core (Caroli and Caldara, 2007). During the last 500 years and until c. 2700 cal BP, the Battaglia basin saw the displacement of mesophilous plants to higher altitudes giving rise to the spread of thermophilous taxa along the coast, and locally, the sheltered bay with a good connection to the sea evolved into a closing lagoon with fresh water input (Caldara et al., 2008).

After the Late Bronze Age crises or collapse, Illyric and Italic peoples settled in Apulia, and the Hellenistic culture spread in southern Italy and the Mediterranean basin giving rise to the Roman civilization in the central regions. Seven sites with archaeobotanical studies testify to the wood exploitation and use of charcoals in Apulia during the Roman and Hellenistic times (e.g. the use for metallurgy in the ancient province of Lecce, Primavera et al., 2012; see in the BRAIN database). Moreover, land-use with the development of wide sylvo-agropastoral systems spread in the regions of southern Italy (Florenzano, 2019).

c. 2240 cal BP (290 BC)

There was a new increase in T CHAC and CHAR with peaks for large microcharcoals and total pollen concentration. The tree percentage decreased, while shrubs increased in a reverse trend with respect to the previous PFAI (recorded at 2660 cal BP). The rise of shrubs was mainly due to Juniperus type and Ephedra, while steppic Artemisia and Chenopodiaceae fell to zero again. Among mountain trees, Fagus fell, while Betula rose, and Pinus tended to expand. Among the mesophilous trees, Acer and Ulmus increased, and deciduous Quercus remains rather steady in the oak woods; Quercus ilex type decreases, and Juniperus type, Olea, and Phillyrea increased after this period. The OJC curve fell to zero again; cereals had a slight decrease, while Cichorieae increased.

Inference

Fires occurred in a phase of relatively reduced biomass and forest cover close to the coast. The fires involved holly oak woods stimulating the spread of shrublands (including macchia), and there were signs of human activities, which were mainly recorded as pastoral land-use. A cool and wet phase was marked by the third minimum of G. sacculifer in the Adriatic cores at 2400–2200 cal BP (Piva et al., 2008). This PFAI occurred during Roman times. The beginning of the continuous curve of OJC, meaning a full development of the cultural landscape, was dated about 100 years later, at c. 2140 cal BP (Figure 6; Mercuri et al., 2013b).

c. 2030 cal BP

A minor rise in T CHAC and CHAR was observed, including a sharp peak of the large size, matching a decrement of total pollen concentration. A spot decrease in tree percentage corresponded to a short increase in shrubs, and this, for example, was represented by the spread of Juniperus type at the expense of Quercus ilex type in the macchia and evergreen woods (Figure 4). In addition, Fagus decreased to zero, while Pinus curve tended to slightly increase, and also Acer increased among the deciduous oak wood. No significant trends were observed in cereal or pastureland taxa.

Inference

The fires would have occurred during a phase of low biomass and forest cover availability, probably close to the coast, and the fire seems to have involved beech and evergreen oaks. Weak signs of human activities were present.

c. 1930 cal BP (AD 20)

A significant presence of large microcharcoals was not concomitant with a significant increase in T CHAC or CHAR, but there was a low peak of total pollen concentration. The tree percentage, already increased, was steady with some increase of shrubs. Fagus and Pinus diminished, while Abies and Picea increased, with traces of Taxus. Deciduous Quercus slightly decreased giving space to Acer, Corylus, and Fraxinus among the mesophilous trees. The Alnus and Salix increase suggested an expansion of riverine woods. Quercus ilex type and Olea were steady, but would increase after this period, while Juniperus type had a short increase. Cereals expanded while Cichorieae gently decreased.

Inference

The fires occurred during a phase of good biomass and forest cover availability, probably close to the coast. Fires seemed to have involved deciduous oaks, pine, and beech. Signs of human activities were present.

c.1510 cal BP (AD 440)

A minor peak of large microcharcoals was not accompanied by an increase of T CHAC and total pollen concentration. The tree percentage was also very high, while the shrub percentage decreased. Mountain trees did not show significant variations, where Juglans had its earlier peak, deciduous Quercus notably increased, while Corylus and Ulmus and Alnus experienced a peak. After a phase of expansion of Quercus ilex type, Juniperus type and Olea, they decreased at this point, while Pistacia cf. lentiscus increased. Cereals were steady and Cichorieae slightly decreased.

Inference

Fires occurred in a phase of high biomass and forest cover, close to the coast. Fires probably involved open areas and macchia, and holly oak among the woods. Signs of human activities were present.

The last millennium PFAIs (from c. 900–865 cal BP to c. 96 cal BP)

c. 900–865 cal BP (AD 1050–1085)

From this date to the present, there were periodic oscillations in the charcoal curves matching several PFAIs that were enhanced in the CHAR record (Figure 2) and commonly occurring in the vicinity along the coast (Figure 3). Some alternations of Fagus and Abies, with Pinus, mirrored the dynamics of mountain/hilly woods; deciduous Quercus is decreasing, and Q. ilex type left space to the macchia with Phillyrea. The OJC curve, including the cultivated Olea, peaked soon after, at 700 cal BP (Mercuri et al., 2013b). The combined increase of cereals and Cichorieae matched the expansion of soil surface devoted to agrarian activities.

c. 530 cal BP (AD 1420)

This occurred after a CHAR increase, and a significant peak of large microcharcoals at c. 570 cal BP, an event with high biomass and peaks of many trees that probably affected the near coast, during a warm phase marked by the last occurrence of G. sacculifer (recorded at c. 550 yr BP, approximately the base of the ‘Little Ice Age’ or LIA; Piva et al., 2008). Fires produced charcoal signals in the phase 500–390 cal BP, without large microcharcoals, which then increased again in one hundred years, at c. 290 cal BP.

c. 120 cal BP (AD 1830)

This was the last sensible peak of CHAR and T CHAC without large microcharcoals. The low values of Fagus, Abies, and Pinus contrasted with the increase in Castanea. Low cereals and Cichorieae were observed but the general impacting presence of people was obvious. This was close to the c. 96 cal BP (AD 1864) peak, which occurred during a cold and humid phase. In fact, two peaks of the benthic foraminifer V. complanata were dated to AD 1689 and 1865. Piva et al. (2008) observed that the ages of these two peaks match the coldest phases of the LIA, the Fernau (AD 1590–1630) and the Napoleon (AD 1810–1820) intervals. These last two episodes recorded in the core RF93-30 occurred quite far from the coast during the Modern Age.

The contribution of fire history to central Mediterranean landscape evolution

Syntheses of both fire and vegetation history have shown that, in Europe, there were uniform patterns and low fire activity during the late Glacial period, while increased frequency and regional variability of the biomass burnt were observed at the early Holocene (Vannière et al., 2016).

In the East Mediterranean, fire frequency and magnitude increased during wetter climatic phases and climatically induced variation in biomass availability that controlled the timing of regional fire activity during pre-Holocene, and again during middle-Holocene times. Then, wildfire cycles appeared to have become more frequent showing changes in the land cover and fuel load availability more reasonably linked to human activity (Turner et al., 2008).

Recently, sedimentary macrocharcoal (>150 µm)-based records from terrestrial cores have demonstrated that pronounced land use changes occurred at a regional scale in central-southern Europe at the beginning of the Neolithic (8–6 ka), during the Bronze Age (5–4 ka), and the Medieval period (1 ka). Specifically, the Mediterranean fire regimes, such as the landscape transformations, have been largely affected by human activities during subsequent cultural phases (Beffa et al., 2016; Mercuri and Sadori, 2014; Tinner et al., 2005, 2016; Vannière et al., 2016). Fire was used to open landscapes also causing the expansion of macchia vegetation in the Mediterranean coastal regions (Noti et al., 2009). Our RF93-30 marine core fell entirely within this period, characterized by the abundance of archaeological sites offering hundreds of archaeobotanical studies from the Italian peninsula (Mercuri et al., 2015a).

Nevertheless, the two oldest PFAIs in this marine core showed quite different signs. The most ancient (7th millennium BP) dated back to the recent Neolithic in Apulia, and most of the fires that contributed to the charcoal record may be reasonably attributed to human activities. This southern Italy region is known to have provided wood supply to humans previously, since Paleolithic people exploited the wood of pines, juniper, pistachio, and rosaceans, and produced charcoals at archaeological sites at least from the Mousterian to Mesolithic phases (>34,000–9000 cal BP at Grotta della Mura; Fiorentino and Parra, 2014).

The PFAI fell after the 6.0 ka climate cooling and retreated southward of the Northern Hemisphere monsoon systems (Steig, 1999), a climatic change causing a rainfall decline or at least a shift toward aridity, and the definitive end of the early Holocene wetter period at c. 6000–5000 cal BP, as seen in the stratigraphical records from central and southern Italy terrestrial archives (Giraudi et al., 2011). In Tuscany, Lago di Massaciuccoli saw a shift toward brackish conditions testifying to a drier environment peaking at 5500 cal BP (Marchetto et al., 2008), while the Tosco-Emilian Apennines forest opening started later (at c. 5000 cal BP, Torbiera di Pavullo: Vescovi et al., 2010). Nevertheless, and in contrast, at about 5500 cal BP, the Adriatic Sea level reached its modern high stand and a circulation pattern similar to the modern one was established with significant environmental changes recorded in the core sediments, and a high sediment supply from the Po River (Piva et al., 2008). Therefore, a wet phase within a dryness trend should have occurred at this period, and an environmental-driven cause could be hypothesized concurring with the 5430 cal BP charcoal signal in the core RF93-30. For the Mediterranean regions, the periodic droughts should have promoted fire activity in the past as it occurs today (Turco et al., 2017).

A further aspect can support the presence of environmental-driven causes at this phase. Reduced vegetation productivity (visible as a decrease of pollen concentration in our core: Figures 6 and 7) was observed after tephra falls that also lead to changes in the forest composition, with Cupressaceae (Juniperus type) being particularly sensitive to tephra deposition in southern Italy (Allen and Huntley, 2018). Tephrostratigraphical correlation of the Adriatic sequences and the Monticchio record for matching key volcanic events revealed the presence of many tephra layers in several central Adriatic cores, including the Agnano Monte Spina-AMST in the core RF95-11 (located NE away from the coast compared with core RF93-30; Figure 7 in Lowe et al., 2007). This was in agreement with the more common east-northeast autumn–winter dispersion of pyroclastic deposits from the Somma-Vesuvius (Rolandi et al., 2007). The AMST predates the AT and have an estimated age ranging from c. 5000 cal yr BP to c. 4300 cal yr BP (Zanchetta et al., 2012), with several eruptions in the Agnano-San Vito area for a few centuries before the main AMS eruption (Iovine et al., 2017). In our record, AT was identified at 529 cm, i.e. 43 cm after the high microcharcoal peak recorded at 572 cm, but the date of this sample (5430 cal BP) seemed to precede the AMST event by several years.

The RF93-30 charcoal record suggested a southwest to northeast direction of winds during the early middle–Holocene phases. Winds, like Sirocco, the hot and dry wind flowing from the south, may have been among the causes of the charcoal transportation to our marine record during the most ancient phases. Remarkably, the summer wind speed is considered to be especially important for fires by Conedera et al. (2018) who discriminate three main fire regime clusters based on the detection of their main climatic, environmental, and socioeconomic drivers in areas with Mediterranean conditions. Today, the presence of winds occurring during dry weather conditions, even in winters (e.g. the Mistral, from the north) makes some regions with Mediterranean vegetation, rich in pines and other sclerophyllous, particularly prone to rapid spreading of fires during summer (like the Maritime Alps; Fréjaville and Curt, 2015).

After the 6.0 ka climate cooling, in Tuscany, the climate change triggered some modification in hydrology, as well as fire regime increase, with tree decline largely involving drought-sensitive trees such as Abies alba (Lago di Massaciuccoli: Colombaroli et al., 2007; Lago dell’Accesa: Drescher-Schneider et al., 2007). In the core RF93-30, a slight Abies decline (and fall of Picea) started giving space to Fagus between c. 4800 and 4200 cal BP. Then, Fagus decline alternated with Abies in a period that was characterized by a climate trend toward dryness (Mercuri et al., 2012; Figure 4). The dry climate was confirmed, for example, by a decrease in the water level at Mezzano (Giraudi, 2004; Sadori, 2018), and studies on climate oscillations in central Italy (Giraudi, 2011). This might be in contrast with the increase of fire frequency observed in many records, including the RF93-30 core, because a dry climate is thought to have been somewhat unfavorable to biomass fuel development. The largest fires occur, in fact, when the precipitation amount in a given area is sufficiently high to allow the full development of vegetation, from the Mediterranean coast to mountain belts (Turco et al., 2017). However, for example, in central Sicily, as in the southern regions, where the vegetation is better adapted to water scarcity, the role of climate forcing on wildfires seems to have been mainly related to regional forest cover type and, hence, to biomass and fuel availability (Sadori et al., 2015). Moreover, human actions altered the effects of the climate in the middle and late Holocene.

The anthropogenic causation of fires may have been increasing during the last four or three millennia, as shown by rather similar evidence to that observed in the French Pyrenees (Rius et al., 2011). From the 4.2 ka, it became hard to disentangle climate and anthropic causation of fire events as both dryness and wood exploitation acted as concomitant forces in reducing trees (Mercuri et al., 2011; Sadori, 2018), also producing evidence of the spreading of sylvo-agropastoral systems in off-site pollen diagrams distant from the area of influence (Mercuri, 2014; Mercuri et al., 2019).

Contrary to the decline of evergreen oaks recorded in South Italy at 4200 cal BP, an important increase of holly oak communities was instead recorded at Massaciuccoli, where regional fire activity had a minimum at c. 4500–3000 cal BP with no clear signs of human activity (Colombaroli et al., 2007). In the Tosco-Emilian Apennines, however, an increase of Cerealia type and other APIs was accompanied by the regional fire activity that peaked at around 4200 cal BP (Pavullo: Vescovi et al., 2010). At Mistras Lagoon (Sardinia), fire activity increased at c. 4100 cal BP in a way such that sclerophyllous shrubland spread, and this was similarly interpreted as being caused by the pressure of Early Bronze Age people on forest communities.

Our data partly agreed with the general fire activity recorded by Vannière et al. (2011) in the regions above 40°N latitude, where transient fire declines were observed at around 6600, 5900, 4200 and 2800 cal BP. Therefore, it is not straightforward that the fire dynamics recorded in our marine core for this range of time may have been exclusively dependent on the human impact. Indeed, the T CHAC/CHAR PFAIs of core RF93-30 at c. 3760 cal BP and c. 3400–2970 cal BP testify to continuous fire activity that could be partly explained by the development of the abovementioned sylvo-agropastoral system of the Terramare culture in the main sedimentary basin of RF93-30, the Po Plain. It is known, in fact, that during the Middle to Recent Bronze Age (3650–3170 BP; Cardarelli, 2009, 2010), this sophisticated culture used fires to open spaces for villages and fields (Montale, Baggiovara: Mercuri et al., 2015c; Poviglio: Cremaschi et al., 2016). Fires were probably employed in the slash and burn of cereal fields, and burnt oaks were found as charred trunks visible at the bottom floor of the Terramare di Montale (Mercuri et al., 2006a, 2006b). Similar Bronze Age fire activity increase has been found around and after 3500 cal BP at Valle di Castiglione (Alessio et al., 1996) and Lago di Mezzano (Sadori, 2018). In Pergusa, fires were seen to have been quite rare with well-established evergreen vegetation during the middle Holocene. Some anthropogenic fires at the beginning of the Bronze Age (around 4000–3500 cal BP) with decreased fires favored the expansion of Olea and Quercus ilex type in central (3200–2500 cal BP; Sadori et al., 2013b) and coastal Sicily (Biviere di Gela, Noti et al., 2009).

Generally speaking, the archaeological layers were especially marked by anthropogenic activities, including the use of fire as fireplaces for culinary or ritual practices, furnaces for metallurgy, slash and burn practices, or forest burning to give space to fields and houses. These fires are evident in macrocharcoal pieces and microcharcoals spread as ash near and at the site. Being rooted in human behavior, the study of charcoals does not have significant chronological limits, and besides the already-mentioned Bronze Age (Fiorentino, 2005), there are many examples of wood uses from the Iron Age (e.g. Castiglioni et al., 2014) and the Middle Ages (Buonincontri et al., 2017) nor geographical diversity (e.g. Turkey, Arslantepe: Masi et al., 2013; Libya, Fewet: Zerboni et al., 2017). In late-Holocene sedimentary contexts, microcharcoal presence and frequency suggest that fire was often intentionally employed for developing sylvo-agropastoral systems with the main examples coming from islands such as Crete over the last 2000 years (Jouffroy-Bapicot et al., 2016). Similar evidence was actually inferred from microcharcoals of lake sediments of Sicily (Pergusa: Sadori et al., 2008) and macrocharcoals from the archaeological sites of the Balearic Islands (Picornell-Gelabert and Servera-Vives, 2017).

In central Italy (Lago Lungo), a fairly significant episode of fire seems to have occurred during the Republican Roman period, at c. 2240 cal BP (290 BC), including some large microcharcoals > 125 µm, and a fire increase was recorded at c. 2200 cal BP at Lago di Massaciuccoli (Colombaroli et al., 2007). Then, the expansion of Mediterranean vegetation with evergreen woods (Quercus ilex type) occurred during Roman times, just prior to the late Antique (c. AD 5th–6th century) when the general reduction of forest could be mainly attributed to the human population increase and impact (Piovesan et al., 2018).

Conclusion

The ubiquitous presence of microcharcoals in the Adriatic Sea core RF93-30 provided clear evidence of the continuous presence of fires in terrestrial vegetation dynamics over the last 7000 years, and it confirmed that significant fire activity could be intercepted by marine cores. Data show that several major fires (related to the combined evidence of high values of microcharcoals and large microcharcoals) occurred from the middle Holocene to Modern Age. The two oldest PFAIs in the charcoal diagram, at c. 6730 cal BP and c. 5430 cal BP, match the contribution of charcoal with the highest values of AP in the sedimentary marine record (Figures 3 and 4). Although these peaks do not represent a single fire event (Whitlock and Larson 2001), our record suggests good correspondence between paleofire activity and terrestrial vegetation biomass during this early phase. Moreover, there is evidence that some fires occurred close to the coast during the most ancient phase, possibly also as an effect of Neolithic peoples’ actions especially affecting pines and mesophilous woods. Indeed, different types of mixed woods were reduced one millennium later by wildfires. A less evident fire activity increase, observed at c. 4150 cal BP, was fairly coincident with a large-scale drought event during which the holly oak woods were probably largely burned in the southern regions. Then, the c. 3760 cal BP fire increase shortly preceded the development of intensive wood exploitation in the Middle Bronze Age cultures. The land use turned toward a continuative plant exploitation that involved fire in territory management practices (Mercuri et al., 2019).

Increases in fire activity were also evident in the fires occurring between the 3rd and 2nd millennia BP (at c. 2660, c. 2240, c. 2030, and c. 1930 cal BP), some of them quite close to the coast, recorded also in the Lago Battaglia record. Finally, a set of fires also became visible in the last millennium (at c. 900–865, c. 530, c. 120–96 cal BP). The marine core RF93-30 shows a trend concordant with the increasing frequency and incidence of human action observed in the Mediterranean basin. A doubling in fire frequency was shown compared with the Holocene average at a regional and continental scale (Vannière et al., 2016). Thus, Vannière and colleagues (2016) stated that ‘no clear distinction could be made between “natural” and “anthropogenic” fire regimes for southern Europe’. Although fire activity has been related to wetter phases and abundant biomass availability as fuel in the past (Daniau et al., 2007), and the biomass burned is known to have decreased from 7 ka due to climate forces (Vannière et al., 2016), technology requests and human impact are known to have improved the pace of fire events and their effects on tree growth. Our record shows that this seems to have involved oak woods. The Mediterranean vegetation, widely adapted to drought, seems to have been able to strongly contrast the negative effects of drought-induced fires. Even today, the fire regime is closely linked to the seasonality of the Mediterranean climate, that is, the dry versus wet seasons allowing an alternation of fuel burning and biomass growth, respectively.

Footnotes

Acknowledgements

Data on archaeobotanical studies are from the BRAIN database and are available at the site . The microcharcoal comparison with the Bronze Age sites of the Po Plain was carried out in the framework of the project ‘SUCCESSO-TERRA – Società Umane, Cambiamenti Climatico-ambientali e Sfruttamento/Sostenibilità delle risorse durante l’Olocene medio in Pianura Padana. Il caso delle Terramare’ fundend by MIUR (Project PRIN 20158KBLNB; PI: Mauro Cremaschi). The authors are very grateful to the anonymous reviewers and the associate editor, Fabienne Marret, for their constructive criticism that let us to improve the text.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Anna Maria Mercuri

Assunta Florenzano

References

Alessio

Allegri

Bella

, et al. (1996) 14C dating, geochemical features, faunistic and pollen analyses of the uppermost 10 m core from Valle di Castiglione (Rome, Italy). Geologica romana 25: 287–308.

Allen

JRM

Huntley

(2018) Effects of tephra falls on vegetation: A Late-Quaternary record from southern Italy. Journal of Ecology 106: 2456–2472.

Beffa

Pedrotta

Colombaroli

, et al. (2016) Vegetation and fire history of coastal north-eastern Sardinia (Italy) under changing Holocene climates and land use. Vegetation History and Archaeobotany 25: 271–289.

Benatti

Bal

Allée

, et al. (2018) Charcoal kilns in the Northern Apennines (Italy): Forest exploitation by past societies in mountain areas. IANSA – Interdisciplinaria Archaeologica 9: 169–178.

Bini

Zanchetta

Persoiu

, et al. (2019) The 4.2 ka BP event in the Mediterranean region: An overview. Climate of the past Discussion 15: 555–577.

Bond

Showers

Cheseby

, et al. (1997) A pervasive millennial-scale cycle in North Atlantic Holocene glacial climates. Science 278: 1257–1266.

Buonincontri

Pecci

Di Pasquale

, et al. (2017) Multiproxy approach to the study of medieval food habits in Tuscany (central Italy). Archaeological and Anthropological Sciences 9: 653–671.

Burjachs

Expósito

(2015) Charcoal and pollen analysis: Examples of Holocene fire dynamics in Mediterranean Iberian Peninsula. Catena 135: 340–349.

Caldara

Caroli

Oronzo

(2008) Holocene evolution sea-level changes in the Battaglia basin area (eastern Gargano coast, Apulia, Italy). Quaternary International 183: 102–114.

10.

Cardarelli

(2009) Insediamenti dell’età del bronzo fra Secchia e Reno. Formazione, affermazione e collasso delle terramare. In: Cardarelli

Malnati

(eds) Atlante dei Beni Archeologici della Provincia di Modena. Collina e Alta Pianura III. Firenze: All’Insegna del Giglio, pp. 33–58.

11.

Cardarelli

(2010) The collapse of the Terramare culture and growth of new economic and social systems during the Late Bronze Age in Italy. In: Cardarelli

Cazzella

Frangipane

, et al. (eds) Atti del Convegno Internazionale ‘Le ragioni del cambiamento/Reasons for change’. Roma: Quaderni di Scienze dell’Antichità, pp. 449–520.

12.

Caroli

(2005) Dinamiche climatiche ed ambientali dell’area garganica nel corso dell’Olocene. PhD Thesis. Bari: Università degli Studi di Bari.

13.

Caroli

Caldara

(2007) Vegetational history of Lago Battaglia (eastern Gargano coast, Apulia, Italy) during the middle–late Holocene. Vegetation History Archaeobotany 16: 317–327.

14.

Carracedo

Cunill

García-Codron

, et al. (2018) History of fires and vegetation since the Neolithic in the Cantabrian Mountains (Spain). Land Degradation & Development 29: 2060–2072.

15.

Castiglioni

Motella

Rottoli

(2014) II combustibile nelle cremazioni dell’Italia nord-orientale. Bulletin de la Société Botanique de France. Actualités Botaniques 139(1992): 311–318.

16.

Clark

(1988) Particle motion and the theory of charcoal analysis: Source area, transport, deposition, and sampling. Quaternary Research 30: 67–80.

17.

Colombaroli

Tinner

Marchetto

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

18.

Colombaroli

Vannière

Chapron

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

19.

Conedera

Krebs

Valese

, et al. (2018) Characterizing Alpine pyrogeography from fire statistics. Applied Geography 98: 87–99.

20.

Cremaschi

Mercuri

Torri

, et al. (2016) Climate change versus land management in the Po Plain (Northern Italy) during the Bronze Age: New insights from the VP/VG sequence of the Terramara Santa Rosa di Poviglio. Quaternary Science Reviews 136: 153–172.

21.

Cui

Marquer

Arzarello

, et al. (2009) An attempt to separate anthropic and natural fire signals in an archaeological context – The case of the Mousterian site Grotta Reali (Rocchetta a Volturno Molise, Central Italy). Frontiers of Earth Science in China 3: 171–174.

22.

Damnati

Reddad

(2017) Les microcharbons dans les sédiments du lac Ifrah moyen atlas (Maroc): significations climatiques depuis 24 000 ans cal. BP. Quaternaire 28: 225–229.

23.

Daniau

A-L

d’Errico

Sánchez-Goñi

, et al. (2010) Testing the hypothesis of fire use for ecosystem management by Neanderthal and Upper Palaeolithic modern human populations. PLoS ONE 5(2): e9157.

24.

Daniau

A-L

Sánchez-Goñi

Beaufort

, et al. (2007) Dansgaard-Oeschger climatic variability revealed by fire emissions in southwestern Iberia. Quaternary Science Reviews 26: 1369–1383.

25.

Di Rita

Magri

(2009) Holocene drought, deforestation and evergreen vegetation development in the central Mediterranean: A 5500 year record from Lago Alimini Piccolo, Apulia, southeast Italy. The Holocene 19: 295–306.

26.

Di Rita

Magri

(2019) The 4.2 ka event in the vegetation record of the central Mediterranean. Climate of the past 15: 237–251.

27.

Di Rita

Melis

(2013) The cultural landscape near the ancient city of Tharros (central West Sardinia): Vegetation changes and human impact. Journal of Archaeological Science 40: 4271–4282.

28.

Di Rita

Simone

Caldara

, et al. (2011) Holocene environmental changes in the coastal Tavoliere Plain (Apulia, southern Italy): A multiproxy approach. Palaeogeography, Palaeoclimatology, Palaeoecology 310: 139–151.

29.

Drescher-Schneider

de Beaulieu

Magny

, et al. (2007) Vegetation history, climate and human impact over the last 15,000 years at Lago dell’Accesa (Tuscany, Central Italy). Vegetation History and Archaeobotany 16: 279–299.

30.

Dupont

Schefuß

(2018) The roles of fire in Holocene ecosystem changes of West Africa. Earth and Planetary Science Letters 481: 255–263.

31.

Fiorentino

(2005) Analisi archeobotanica. In: Martinelli

(ed.) Il Villaggio dell’età del Bronzo Medio di Portella a Salina nelle Isole Eolie. Firenze: Origines, pp. 263–273.

32.

Fiorentino

Magri

(2008) Charcoals from the Past: Cultural and Palaeoenvironmental Implications: Proceedings of the Third International Meeting of Anthracology, Cavallino, Lecce (Italy). Oxford: Archaeopress (BAR International Series 1807).

33.

Fiorentino

Parra

(2014) ‘Lost’ postglacial littoral environments in SE Italy: Anthracological evidence at Grotta delle Mura. Review of Palaeobotany and Palynology 218: 198–203.

34.

Fiorentino

Caldara

De Santis

, et al. (2013) Climate changes and human–environment interactions in the Apulia region of southeastern Italy during the Neolithic period. The Holocene 23: 1297–1316.

35.

Florenzano

(2013) Evolution of a Mediterranean landscape as shown by the archaeo-environmental reconstruction of Lucanian site. PhD Thesis, Università di Modena e Reggio Emilia.

36.

Florenzano

(2019) The history of pastoral activities in S Italy inferred from palynology: A long-term perspective to support biodiversity awareness. Sustainability 11: 404.

37.

Florenzano

Marignani

Rosati

, et al. (2015) Are Cichorieae an indicator of open habitats and pastoralism in current and past vegetation studies? Plant Biosystems 149: 154–165.

38.

Florenzano

Mercuri

Pederzoli

, et al. (2012) The significance of intestinal parasite remains in pollen samples from Mediaeval pits in the Piazza Garibaldi of Parma, Emilia Romagna, Northern Italy. Geoarchaeology 27: 34–47.

39.

Fréjaville

Curt

(2015) Spatiotemporal patterns of changes in fire regime and climate: Defining the pyroclimates of south-eastern France (Mediterranean basin). Climatic Change 129: 239–251.

40.

Giraudi

(2004) Evoluzione tardo-olocenica del delta del Tevere. Il Quaternario-Italian Journal of Quaternary Sciences 17: 477–482.

41.

Giraudi

(2011) Comparison between lake-level changes, pollen and sedimentological data at Lago di Mezzano (VT), central Italy. Alpine and Mediterranean Quaternary 24: 129–131.

42.

Giraudi

Magny

Zanchetta

, et al. (2011) The Holocene climatic evolution of Mediterranean Italy: A review of the continental geological data. The Holocene 21: 105–115.

43.

Iovine

Fedele

Mazzeo

, et al. (2017) Timescales of magmatic processes prior to the ∼4.7 ka Agnano-Monte Spina eruption (Campi Flegrei caldera, Southern Italy) based on diffusion chronometry from sanidine phenocrysts. Bulletin of Volcanology 79: 18.

44.

Jones

Chaloner

Kuhlbusch

TAJ

(1997) Proposed bio-geological and chemical based terminology for fire-altered plant matter. In: Clark

Cachier

Goldammer

, et al. (eds) Sediment Records of Biomass Burning and Global Change. Berlin: Springer, pp. 9–22.

45.

Jouffroy-Bapicot

Vannière

Iglesias

, et al. (2016) 2000 Years of grazing history and the making of the Cretan Mountain Landscape, Greece. PLoS ONE 11(6): e0156875.

46.

Kaltenrieder

Procacci

Vanniere

, et al. (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.

47.

Kaniewski

Marriner

Cheddadi

, et al. (2018) The 4.2 ka BP event in the Levant. Climate of the Past 14: 1529–1542.

48.

Langone

Asioli

Correggiari

, et al. (1996) Age-depth modelling through the late Quaternary deposits of the central Adriatic basin. In: Guilizzoni

Oldfield

(eds) Special volume: Palaeoenvironmental Analysis of Italian Crater Lake and Adriatic Sediments (PALICLAS), vol. 55. Pallanza: Memorie dell’ Istituto Italiano di Idrobiologia, pp. 177–196.

49.

Languut

Cheddadi

Carrión Sebastián

, et al. (2019) The origin of olive domestication in the Mediterranean Basin: The fossil pollen evidence. The Holocene. Epub ahead of print 14 February. DOI: 10.1177/0959683619826654.

50.

Lowe

Accorsi

CAA

Bandini Mazzanti

, et al. (1996) Pollen stratigraphy of sediment sequences from lakes Albano and Nemi (near Rome) and from the central Adriatic, spanning the interval from oxygen isotope Stage 2 to the present day. In: Guilizzoni

Oldfield

(eds) Special volume: Palaeoenvironmental Analysis of Italian Crater Lake and Adriatic Sediments (PALICLAS), vol. 55. Pallanza: Memorie dell’ Istituto Italiano di Idrobiologia, pp. 71–98.

51.

Lowe

Blockley

Trincardi

, et al. (2007) Age modelling of late Quaternary marine sequences in the Adriatic: Towards improved precision and accuracy using volcanic event stratigraphy. Continental Shelf Research 27: 560–582.

52.

Marchetto

Colombaroli

Tinner

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

53.

Marignani

Chiarucci

Sadori

, et al. (2017) Natural and human impact in Mediterranean landscapes: An intriguing puzzle or only a question of time? Plant Biosystems 151: 900–905.

54.

Mariotti Lippi

Florenzano

Rinaldi

, et al. (2018) The Botanical Record of Archaeobotany Italian Network-BRAIN: A cooperative network, database and website. Flora Mediterranea 28: 365–376.

55.

Marquer

(2010) From microcharcoal to macrocharcoal: Reconstruction of the ‘wood charcoal’ signature in Paleolithic archaeological contexts. In: The Taphonomy of Burned Organic Residues and Combustion Features in Archaeological Contexts. Proceedings of the Round Table (eds Théry-Parisot

Chabal

Costamagno

), vol. 2, Valbonne, 27–29 May, pp. 105–115. Toulouse: P@lethnologie.

56.

Masi

Sadori

Zanchetta

, et al. (2013) Climatic interpretation of carbon isotope content of mid-Holocene archaeological charcoals from eastern Anatolia. Quaternary International 303: 64–72.

57.

Mercuri

(2014) Genesis and evolution of the cultural landscape in central Mediterranean: The ‘where, when and how’ through the palynological approach. Landscape Ecology 29: 1799–1810.

58.

Mercuri

Accorsi

Bandini Mazzanti

, et al. (2006a) Cereal fields from the Middle-Recent Bronze Age, as found in the Terramara di Montale, in the Po Plain (Emilia Romagna, Northern Italy), based on pollen, seeds/fruits and microcharcoals. In: Morel

(ed.) The Archaeology of Crop Fields and Gardens. Bari: Edipuglia, pp. 251–270.

59.

Mercuri

Accorsi

Bandini Mazzanti

, et al. (2006b) Economy and environment of Bronze Age settlements-Terramaras-in the Po Plain (Northern Italy): First results of the archaeobotanical research at the Terramara di Montale. Vegetation History and Archaeobotany 16: 43–60.

60.

Mercuri

Allevato

Arobba

, et al. (2015a) Pollen and macroremains from Holocene archaeological sites: A dataset for the understanding of the bio-cultural diversity of the Italian landscape. Review of Palaeobotany and Palynology 218: 250–266.

61.

Mercuri

Sadori

(2014) Mediterranean culture and climatic change: Past patterns and future trends. In: Goffredo

Dubinsky

(eds) The Mediterranean sea. Dordrecht: Springer, pp. 507–527.

62.

Mercuri

Bandini Mazzanti

Florenzano

, et al. (2013a) Anthropogenic pollen indicators (API) from archaeological sites as local evidence of human-induced environments in the Italian peninsula. Annali Di Botanica (Roma) 3: 143–153.

63.

Mercuri

Bandini Mazzanti

Florenzano

, et al. (2013b) Olea, Juglans and Castanea: The OJC group as pollen evidence of the development of human induced environments in the Italian peninsula. Quaternary International 303: 24–42.

64.

Mercuri

Bandini Mazzanti

Torri

, et al. (2012) A marine/terrestrial integration for mid-late Holocene vegetation history and the development of the cultural landscape in the Po Valley as a result of human impact and climate change. Vegetation History and Archaeobotany 21: 353–372.

65.

Mercuri

Florenzano

Burjachs

, et al. (2019) From influence to impact: The multifunctional land-use in Mediterranean prehistory emerging from palynology of archaeological sites (8.0–2.8 ka BP). The Holocene. Epub ahead of print 15 February. DOI: 10.1177/0959683619826631.

66.

Mercuri

Marignani

Sadori

(2015b) Palaeoecology and long-term human impact in plant biology. Plant Biosystems 149: 136–143.

67.

Mercuri

Montecchi

Pellacani

, et al. (2015c) Environment, human impact and the role of trees on the Po plain during the Middle and Recent Bronze Age: Pollen evidence from the local influence of the terramare of Baggiovara and Casinalbo. Review of Palaeobotany and Palynology 218: 231–249.

68.

Mercuri

Sadori

Uzquiano Ollero

(2011) Mediterranean and north-African cultural adaptations to mid-Holocene environmental and climatic changes. The Holocene 21: 189–206.

69.

Millington

JDA

Wainwright

Perry

GLW

, et al. (2009) Modelling Mediterranean landscape succession-disturbance dynamics: A landscape fire-succession model. Environmental Modelling & Software 24: 1196–1208.

70.

Nelle

Robin

Talon

(2013) Pedoanthracology: Analysing soil charcoal to study Holocene palaeoenvironments. Quaternary International 289: 1–4.

71.

Nisbet

(2008) Environment and agriculture in the early Neolithic of the Arene Candide (Liguria). In: Fiorentino

Magri

(eds) Charcoals from the Past: Cultural and Palaeoenvironmental Implications. Proceedings of the Third International Meeting of Anthracology, Cavallino, Lecce (Italy). Oxford: Archaeopress (BAR International Series 1807), pp. 193–198.

72.

Noti

van Leeuwen

JFN

Colombaroli

, et al. (2009) Mid- and late-Holocene vegetation and fire history at Biviere di Gela, a coastal lake in southern Sicily, Italy. Vegetation History and Archaeobotany 18: 371–387.

73.

Oldfield

Asioli

Accorsi

, et al. (2003) A high resolution late Holocene palaeo-environmental record from the central Adriatic Sea. Quaternary Science Reviews 22: 319–342.

74.

Palmisano

Bevan

Shennan

(2017) Comparing archaeological proxies for long-term population patterns: An example from central Italy. Journal of Archaeological Science 87: 59–72.

75.

Peresani

Ravazzi

Pini

, et al. (2018) Human settlement and vegetation-climate relationships in the Greenland Stadial 5 at the Piovesello site (Northern Apennines, Italy). Quaternary Research 90: 503–528.

76.

Picornell-Gelabert

Servera-Vives

(2017) Landscape practices and everyday life in domestic spaces in Bronze Age Mallorca (Balearic Islands): Perspectives for and archaeology of fuel and firewood. Quaternary International 431: 73–89.

77.

Piovesan

Mercuri

Mensing

(2018) The potential of paleoecology for functional forest restoration planning: Lessons from Late Holocene Italian pollen records. Plant Biosystems 152: 508–514.

78.

Piva

Asioli

Trincardi

, et al. (2008) Late-Holocene climate variability in the Adriatic Sea (Central Mediterranean). The Holocene 18: 153–167.

79.

Primavera

Fiorentino

Colaianni

(2012) Il combustibile delle attività metallurgiche nelle forge di Lecce tardo-antica: caratteristiche della vegetazione e sfruttamento dell’ambiente. In: Giardino

(ed.) Archeometallurgia: dalla conoscenza alla fruizione. Bari: Edipuglia, pp. 321–331.

80.

Proske

Wood

Fallon

, et al. (2015) Use of heavy liquid density separation to remove pyrite from sediment samples for radiocarbon dating. Quaternary Geochronology 25: 66–71.

81.

Reddad

Etebaai

Benhardouze

, et al. (2012) Paleofire activity in North Africa during the last 21 ka inferred from a charcoal record from lake Ifrah (Middle atlas – Morocco): Climatic implications. Journal of African Earth Sciences 84: 47–53.

82.

Rius

Vannière

Galop

, et al. (2011) Holocene fire regime changes from multiple-site sedimentary charcoal analyses in the Lourdes basin (Pyrenees, France). Quaternary Science Reviews 30: 1696–1709.

83.

Rolandi

Paone

Di Lascio

, et al. (2007) The 79 AD eruption of Somma: The relationship between the date of the eruption and the southeast tephra dispersion. Journal of Volcanology and Geothermal Research 169: 87–98.

84.

Rottoli

Castiglioni

(2009) Prehistory of plant growing and collecting in Northern Italy, based on seed remains from the Early Neolithic to the Chalcolithic ( c. 5600–2100 cal BC). Vegetation History and Archaeobotany 18(1): 91–103.

85.

Rull

(2010) Ecology and palaeoecology: Two approaches, one objective. The Open Ecology Journal 3: 1–5.

86.

Sadori

(2018) The Lateglacial and Holocene vegetation and climate history of Lago di Mezzano (central Italy). Quaternary Science Reviews 202: 30–44.

87.

Sadori

Giardini

(2007) Charcoal analysis, a method to study vegetation and climate of the Holocene: The case of Lago di Pergusa, Sicily (Italy). Geobios 40: 173–180.

88.

Sadori

Bertini

Combourieu-Nebout

, et al. (2013a) Palynology and Mediterranean vegetation history. Flora Mediterranea 23: 141–156.

89.

Sadori

Masi

Ricotta

(2015) Climate-driven past fires in central Sicily. Plant Biosystems 149: 166–173.

90.

Sadori

Ortu

Peyron

, et al. (2013b) The last 7 millennia of vegetation and climate changes at Lago di Pergusa (central Sicily, Italy). Climate of the past 9: 1969–1984.

91.

Sadori

Zanchetta

Giardini

(2008) Last Glacial to Holocene palaeoenvironmental evolution at Lago di Pergusa (Sicily, Southern Italy) as inferred by pollen, microcharcoal, and stable isotopes. Quaternary International 181: 4–14.

92.

Steig

(1999) Mid-Holocene climate change. Science 286: 1485–1487.

93.

Tinner

(2003) Size parameters, size-class distribution and area-number relationship of microscopic charcoal: Relevance for fire reconstruction. The Holocene 13: 499–505.

94.

Tinner

Conedera

Ammann

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

95.

Tinner

Vescovi

van Leeuwen

JFN

, et al. (2016) Holocene vegetation and fire history of the mountains of Northern Sicily (Italy). Vegetation History and Archaeobotany 25: 499–519.

96.

Trincardi

Cattaneo

Asioli

, et al. (1996) Stratigraphy of the Late-Quaternary deposits in the central Adriatic basin and the record of short-term climatic events. In: Guilizzoni

Oldfield

(eds) Special Volume: Palaeoenvironmental Analysis of Italian Crater Lake and Adriatic Sediments (PALICLAS), vol. 55. Pallanza: Memorie dell’Istituto Italiano di Idrobiologia, pp. 39–70.

97.

Turco

von Hardenberg

AghaKouchak

, et al. (2017) On the key role of droughts in the dynamics of summer fires in Mediterranean Europe. Nature Scientific Reports 7: 81.

98.

Turner

Roberts

Jones

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

99.

Vannière

Blarque

Rius

, et al. (2016) 7000-year human legacy of elevation-dependent European fire regimes. Quaternary Science Reviews 132: 206–212.

100.

Vannière

Colombaroli

Chapron

, et al. (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.

101.

Vannière

Power

Roberts

, et al. (2011) Circum-Mediterranean fire activity and climate changes during the mid-Holocene environmental transition (8500–2500 cal. BP). The Holocene 21: 53–73.

102.

Verardo

(1997) Charcoal analysis in marine sediments. Limnology and Oceanography 42: 192–197.

103.

Vescovi

Ammann

Ravazzi

, et al. (2010) A new Late-glacial and Holocene record of vegetation and fire history from Lago del Greppo, northern Apennines, Italy. Vegetation History and Archaeobotany 19: 219–233.

104.

Weiss

Courty

Wetterstrom

, et al. (1993) The genesis and collapse of third millennium north Mesopotamian civilization. Science 261: 995–100.

105.

Whitlock

(2001) Variations in Holocene fire frequency: A view from the western United States. Biology & Environment Proceedings of the Royal Irish Academy 101B(1): 65–77.

106.

Whitlock

Larson

(2001) Charcoal as a fire proxy. In: Smol

Birks

HJB

Last

(eds) Tracking Environmental Change Using Lake Sediments. Vol. 3: Terrestrial, Algal, and Siliceous Indicators. Dordrecht: Kluwer Academic Publishers, pp. 75–97.

107.

Zanchetta

Giraudi

Sulpizio

, et al. (2012) Constraining the onset of the Holocene ‘Neoglacial’ over the central Italy using tephra layers. Quaternary Research 78: 236–247.

108.

Zerboni

Mori

Bosi

, et al. (2017) Domestic firing activities and fuel consumption in a Saharan oasis: Micromorphological and archaeobotanical evidence from the Garamantian site of Fewet (Central Sahara, SW Libya). Journal of Arid Environments 157: 149–167.