Palaeoecological studies in the northern Apennines (embracing the regions of Liguria, and parts of Tuscany and Emilia-Romagna) have enabled reconstruction of the main patterns of environmental change from Late Würm Late Glacial to the end of the middle Holocene. The data generated by the palaeoecological study of the upland mire of Rovegno (812m asl; Liguria) indicates that prior to 17,056-16,621 cal yrs BP until 15,099-14,501 cal yrs BP sediment was deposited in a freshwater lake and sedge-swamp with dryland vegetation characterised by open Pinus woodland, Artemisia dominated grassland, Juniperus shrubland within an arctic-tundra, steppe or semi-desert environment during the late Würm glaciation. The presence of Pinus woodland prior to 17,056-16,621 cal yrs BP at Rovegno is particularly interesting and suggests the onset of 'warmer' climatic conditions following the last glacial maximum. In contrast, the decline of Pinus at approximately 17,056-16,621 cal yrs BP suggests the renewal of colder conditions, and may be provisionally equated with the North Atlantic Heinrich 1 (H1) rapid climate change event. From 15,099-14,501 cal yrs BP to the onset of the Holocene at ~11,700 cal yrs BP was characterised by the establishment of Betula and Pinus woodland, and the expansion of thermophilous taxa, namely Abies, Corylus and Quercus, during the Late Würm Lateglacial Interstadial. The paucity of Abies in the Lateglacial pollen data from Rovegno is perhaps the most important contrast with other sites, and may be explained by differences in elevation and proximity of glacial refugia. The decline of Pinus woodland sometime after 13,387-13,163 cal yrs BP, and increase in Betula and Artemisia, at Rovegno may possibly be equated with the Late Würm Lateglacial Stadial ('Younger Dryas'). The absence of a clear expansion of herbaceous taxa in the pollen record, however, suggests that the event is poorly represented. In contrast, at other sites, such as Prato Spilla 'C', the event is clearly marked by the decline of Quercus woodland, and the expansion of herbaceous taxa. The return of colder climatic conditions at this time has been recorded in several lake sequences in the Mediterranean and Italy. During the early Holocene, the dryland vegetation comprised mixed coniferous and deciduous woodland dominated by Quercus with Abies, Fagus and Corylus. Fagus woodland becomes established at Rovegno sometime before 6488-6318 cal yrs BP, but never becomes a major component of the woodland cover. At higher elevation, mixed deciduous -- coniferous woodland, comprising Abies with Pinus, Betula, Fraxinus and Quercus dominated. At Rovegno, the 'abrupt' nature of the 'Corylus rise' suggests that it occurred following a hiatus in sedimentation during the earliest Holocene. The timing and duration of the hiatus remains uncertain. Although the cause of the 'Corylus rise' has been attributed to climate change, distance from glacial refugia, and the migration rate of Corylus, competition or Mesolithic human activities, the data from Rovegno add very little to the current debate. For the remainder of the early Holocene, Abies dominated the upland zone of the Northern Apennines, forming a mixed coniferous-deciduous woodland cover with Quercus, Tilia, Ulmus and Corylus, whilst at Rovegno there was a higher proportion of Quercus within the Abies woodland and the increasing presence of Fagus. The onset of the Middle Holocene was marked by contrasting vegetation patterns in the upland zone of the Northern Apennines: (a) a sustained reduction in Abies woodland, (b) a temporary decline in Abies, (c) the persistence of Abies, and (d) the broadly synchronous expansion of Fagus woodland. This general pattern coincided with increases in Quercus and Corylus, and herbaceous taxa (e.g. Plantago lanceolata), and a reduction in Tilia, Ulmus and Fraxinus. At Rovegno, Fagus woodland was already well established, which is consistent with many lower elevation sites, whilst Quercus became increasingly dominant over Abies. The presence of Fagus from 6488-6318 cal yrs BP at Rovegno confirms that this taxon colonised mid-altitudes prior to its major expansion at higher elevations from ~6000 cal yrs BP. From ~6000 cal yrs BP, Fagus formed mixed Abies-Fagus woodland across the upland zone of the Northern Apennines, which coincided with further reductions in Ulmus, Tilia and Fraxinus, increases in herbaceous taxa (e.g. Plantago lanceolata and Poaceae), as well as Carpinus and Ostrya. At lower elevations, such as Rovegno, Quercus became dominant, with Fagus only forming a minor component of the woodland cover. The period from ~4700 cal yrs BP to the end of the middle Holocene (~3000 cal yrs BP) records a further major change in vegetation succession with the beginning of a sustained reduction in Abies woodland, and an overall increase in Fagus, which suggests dynamic changes in the composition and structure of the woodland cover. The decline of the main woodland taxa at many sites was accompanied by occasional evidence for soil erosion and a general increase in non-arboreal taxa; this commenced during the Early and Middle Neolithic (~7800-6100 cal yrs BP), and continued during the Late Neolithic (~6100-5800 cal yrs BP), Copper Age (~5800-4200 cal yrs BP) and Bronze Age (~4200-2900 cal yrs BP). Although these changes are often associated with the spread of agriculture in many parts of Europe, the palaeoecological evidence for human activities at mid and high elevations in the Northern Apennines remains equivocal. Therefore, the application of standard criteria for inferring human impact on the environment using palaeoecological data, such as woodland clearance, biomass burning and soil erosion remain uncertain. Additional methods must therefore be routinely employed in northern Apennine research to avoid ongoing circular arguments, such as non-pollen palynomorph (NPP) analysis, especially coprophilous fungal spores. At Prato Spilla 'A' (Emilia-Romagna), previous pollen analysis has inferred two episodes of human interference in vegetation succession during the Middle and Late Neolithic. The low values of coprophilous fungal spores across both of these events suggests, however, that there is little or no evidence for human activity associated with pastoralism, and instead the dung spores could be ascribed to wild mountain herbivores, most likely ungulates and lagomorphs. The fungal data suggest, therefore, that greater caution must be used when interpreting human activity based upon pollen data alone. [ABSTRACT FROM AUTHOR]