The distribution of palaeoenvironmental evidence for landscape and land use dynamics in the area is patchy in space and time. Detailed insights into land use history are available for a few locations and periods. These are best viewed as vignettes which can be used to derive hypotheses about the pattern and nature of human use of natural resources, rather than generalisations. This is most detailed for Glen Shee where work focused on the first millennium AD (Strachan et al 2019), and Loch Tay which focused on around the last 400 years and the full Holocene (Hamilton et al 2009; Atkinson 2016). In the latter case, it is unfortunate that the palynological investigations associated with excavations on the north shore remain almost entirely in the grey literature (Tipping et al 2009a; 2016). In contrast, the broad sweep of glacial and Holocene vegetation history can more readily be set into wider context through reference to regional glacial and biogeographical trends (Tipping 1994). The region has to some extent fallen between the gaps in previous regional reviews of vegetation history, such as those published in a special issue of the Botanical Journal of Scotland in 1997. This may reflect the ecotonal location of the area, which spans the central-eastern lowlands and Highland fringe.
Pollen analysis is the most commonly used palaeoenvironmental and -ecological proxy throughout the Holocene. Most pollen records derive from peat; studies of lake sediments are limited (eg Caseldine 1980; Milburn 1997), with a cluster from Loch Leven, where the diatoms and macroscopic plant remains detail the impacts of changing land use over the last 200 years (eg Salgado et al 2010; Bennion et al 2012). Plant macrofossils, carbonised or waterlogged, are most frequent in on-site contexts, while sources like soil micromorphology (eg McKenzie and Simpson in Atkinson 2016) and dendrochronology and -ecology are scarce (Crone 2014). More recent developments, like biomolecular techniques such as lipids, bone isotopes, sedimentary or ancient DNA (eg Gigleux et al 2017; Mackay et al 2020) and non-pollen palynomorphs (eg dung fungi; Farrell 2015; Davies 2019), appear to be absent. Wider review of the literature on these proxies is required to evaluate the scope for future research to add detail, depth and colour to the archaeological heritage of this region.
As indicated above, this review focuses on the off-site, landscape record. It does not consider climate change or on-site sampling of environmental remains in any depth. On-site plant macrofossil and wood remains offer valuable additional insights into vegetation diversity and species associations to complement and extend pollen stratigraphic sequences. This is because many species found on-site may be under-represented or absent due to poor pollen dispersal or taxonomic limitations on identification. For example, insect-pollinated hawthorn, rowan, bramble and raspberry all belong to the Rosaceae family and are difficult to identify to genus or species level from their pollen, but may be identified through seed or wood remains. Archaeobotanical samples can also offer closer insight into husbandry and processing techniques based on the remains of crops and their weed assemblages (eg Miller and Ramsay in Atkinson 2016). Soil sampling for micromorphology or pollen has been applied in few contexts (eg Caseldine 1983; Carter in Rideout et al 1996; Simpson and Davidson 1998; Simpson and McKenzie in Atkinson 2016). These on-site contexts are compared with off-site pollen records to help understand spatial variability in past landscapes.
Climate change is best reconstructed through climate-sensitive proxies (eg chironomids, beetles, dendroclimatology, lake or peatland watertables). While some records from this area provide potential contributions to broader reconstructions of climate change, no studies in the Perth and Kinross area focus specifically on climate reconstruction. The scale of PKARF precludes a review of regional to national level palaeoclimate evidence. The importance of high quality palaeoclimate data, that is, high temporal resolution with clarity over which elements of climate are being reconstructed (eg temperature, precipitation, seasonality), cannot be underestimated, however. Inferences of settlement marginality in the archaeological record may be overly influenced by modern perceptions and reference to general or outdated climate data. Studies focused on sea level and glacial change (eg Cullingford et al 1980) are also outwith the scope of this review.
Research Priorities across all periods
Improved understanding and comparison of upland/lowland environments and land–use dynamics through time. Although the topography of Perth and Kinross is characterised by Highland/Lowland divide, most studies have focused either on lower altitude sites and settings (eg Hulme and Shirriffs 1986; Milburn 1997) or higher ground (eg Tipping et al 1993; Tipping 1995) rather than working across altitudinal gradients. Two exceptions are the PhD research of Caseldine (1980) in eastern Perthshire, which lacks a secure chronology and the loch shore to montane distribution of sites in the Ben Lawers project (Atkinson 2016; Tipping et al 2009a). With these exceptions, there is currently limited integration of palaeoecological records to assess spatio-temporal differences in vegetation composition and human activity across low/high ground. As Dalglish (2012) comments, this can lead to an overly simplistic separation of highland and lowland zones which reflects value judgements of the authors more than the complexity and diversity of life and agriculture in each zone. There is clear opportunity and value in addressing this gap in future work.