Criteria for separation of Erica tetralix and E. ciliaris-type pollen (represented here by E. vagans). The measurements shown are accurate for the grains pictured, which were prepared using standard procedures and mounted in silicon oil.

... Separation of ericaceous species was an important aspect of this project owing to the need to distinguish between the pollen of Erica tetralix (cross-leaved heath) and the much rarer species, Erica vagans (Cornish heath). Based on the key of Moore et al (1991) and careful examination of examples in the reference collection, it proved possible to ascribe many of the Erica grains encountered to either E. tetralix or E. ciliaris-type, which includes a number of species, including Erica ciliaris (Dorset Heath) and E. vagans. Grains were examined at x1000 magnification and typed according to the criteria shown in Figure 3. Where features were indistinct or pollen was obscured, broken or crushed, it was often impossible to determine the species/type. In these cases grains were defined as a) Erica undiff. (undifferentiated) for those definitely of Erica, or b) Ericaceae undiff., for grains that could not be identified as Erica rather than Calluna with certainty. Radiocarbon dates for the basal part of the Lower Lancarrow sediments indicate that they accumulated in the late-glacial period. Consequently, it was thought likely that some of the Betula (birch) pollen in the assemblage originated from dwarf birch (Betula nana) rather than tree birches (B. pubescens or B. pendula in this area (Huntley and Birks 1983)). Pollen grains of B. pendula (silver birch), B. pubescens (downy birch) and B. nana cannot be distinguished easily at x400 magnification and require careful measurement at x1000 with oil immersion. Generally, tree birch grains are larger than those of dwarf birch and have deeper pores relative to their size. There have been numerous studies on the separation of Betula pollen types using measurements of grain diameter (D), pore depth (P) (Figure 4) and the ratio between these (D/P), based on measurements of modern birch pollen (eg Birks 1968b;Prentice 1981;M?kel? 1996;1998;Caseldine 2001;Clegg et al 2005;Karlsd?ttir et al 2007;Karlsd?ttir et al 2008). Some pollen falls between the types in each of the studies, and the number of grains assigned to a species depends to some extent on the method employed. All Betula grains encountered in samples from Lower Lancarrow were photographed using a Leica DFC320 camera (mounted on the microscope) and measured with calibrated imaging software (Leica Application Suite). D and P were recorded for each pore and side of the grain wherever possible, in order to assess variation within individual grains. Deciding on a method to differentiate the birch pollen was difficult; Birks (1968b) found that D/P was the best means of distinguishing B. nana from B. pubescens, while Karlsd?ttir et al (2007) found that grain diameter (D) gave a clearer separation with fewer grains in the overlap zone. M?kel? (1996) investigated the differences between B. nana, B. pubescens and B. pendula, finding that mean D for each species is distinct (Table 1), whereas Prentice et al (1981) report significant overlaps in all of the measurements. B. pendula is intermediate between B. nana and B. pubescens in terms of D and P; as all three types are native to the UK and can occur on heath/moorland (see Clapham et al 1962; Stace 1997) it is not possible to exclude any of them with confidence. Table 1: Mean values of D (diameter -see Figure 4) for Betula pubescens, B. pendula and B. nana (after M?kel? ...