Salix transect of Europe: variation in ploidy and genome size in willow-associated common nettle, Urtica dioica L. sens. lat., from Greece to arctic Norway

Abstract Background The common stinging nettle, Urtica dioica L. sensu lato, is an invertebrate "superhost", its clonal patches maintaining large populations of insects and molluscs. It is extremely widespread in Europe and highly variable, and two ploidy levels (diploid and tetraploid) are known. However, geographical patterns in cytotype variation require further study. New information We assembled a collection of nettles in conjunction with a transect of Europe from the Aegean to Arctic Norway (primarily conducted to examine the diversity of Salix and Salix-associated insects). Using flow cytometry to measure genome size, our sample of 29 plants reveals 5 diploids and 24 tetraploids. Two diploids were found in SE Europe (Bulgaria and Romania) and three diploids in S. Finland. More detailed cytotype surveys in these regions are suggested. The tetraploid genome size (2C value) varied between accessions from 2.36 to 2.59 pg. The diploids varied from 1.31 to 1.35 pg per 2C nucleus, equivalent to a haploid genome size of c. 650 Mbp. Within the tetraploids, we find that the most northerly samples (from N. Finland and arctic Norway) have a generally higher genome size. This is possibly indicative of a distinct population in this region.


Introduction
During a recent study of willow (Salix spp.) stands on a latitudinal transect across Europe (Cronk et al. 2015) the opportunity arose to sample individuals of Urtica dioica L. ssp. dioica (the common stinging nettle) that frequently co-occurs with willow in riparian habitats (see under Materials and Methods for further details). Urtica dioica is one of the most remarkable plants of Europe. First it possesses a defense, stinging hairs, which are a small marvel of biochemistry and biomechanics. These are highly effective against vertebrate herbivores (Levin 1973, Pollard and Briggs 1984, Pullin and Gilbert 1989, Tuberville et al. 1996. The cell walls of the trichome tip are silicified and brittle (Haberlandt 1914, Barber and Shone 1966, Thurston and Lersten 1969, Thurston 1974, Sowers and Thurston 1979 and break off (like the tip of a glass ampoule) on the slightest mechanical stimulation. The fluid released is a potent and complex mixture of toxins including histamine, oxalic acid and tartaric acid (Emmelin and Feldberg 1949, Fu et al. 2006, Taskila et al. 2000.
Secondly it has an extraordinary biogeographical range, occurring in every corner of Europe, from the shores of the Mediterranean to the Arctic Ocean and from the winter-cold central European plain to the rainswept coasts of western Ireland. Few plants have the ability to grow in such a wide range of climatic conditions. Over this range it is largely native, having spread along its natural habitat of rich alluvial river floodplains. However, it has also become an aggressive ruderal, taking advantage of human disturbance to complete its conquest of Europe through accidental introduction by humans.
Thirdly it is an invertebrate "super-host". Throughout Europe it provides the food plant for large numbers of specialist and generalist insects, notably in the Lepidoptera, Coleoptera and Hemiptera (Davis 1973, Davis 1975, Davis 1983, Davis 1989, Perrin 1975. Fourthly, it has exceptional mineral nutrition, being highly phosphate demanding. It ceases growth if phosphate is limiting and responds luxuriantly if phosphate is added, whereas in contrast plants adapted to poor soil scarcely respond to additional phosphate (Pigott 1971, Pigott and Taylor 1964, Taylor 2009). It is not only an indicator of high available phosphate, but it is also a general mineral accumulator, having high concentrations of calcium, nitrogen and phosphorus in its tissues (Müllerová et al. 2014). This may go some way to explaining its attractiveness to herbivorous invertebrates. U. membranacea Poir. 2n=22 (Corsi et al. 1999) Two types of cytological diversity have been found in Urtica dioica subsp. dioica. One is the reported difference in tetraploid chromosome number between 2n=48 and 2n=52 (Skalińska et al. 1974). Such a discrepancy could be due to miscounts, but the repeated reports of both numbers leads to a suspicion that both numbers do exist in nature.
There is also the difference in ploidy level. The possibility must be entertained that counts for Urtica dioica of 2n=26 (diploid) refer to one of the infraspecifc taxa and not to U. dioica

Context of study
Urtica dioica samples (  Table 2.
Previous genome size estimates in Urtica dioica s.l.
*Fe = Feulgen microdensitometry, FC:PI = Flow cytometry using propidium iodide of this was to survey variation in Urtica and one of its constant herbivores, the Urtica psyllid, Trioza urticae, which were co-sampled. Information on Trioza will be the given in separate papers. Herbarium and living Urtica samples were collected. The aim of the current work was to investigate the extent of ploidy level and genome diversity within the resultant Urtica collection.

Site selection and sampling
Full details of the sites (mainly riverine alluvial habitats), and their selection are given in Cronk et al. (2015). The sites are summarized in Table 3. In all, 42 Salix sites were chosen between Athens (Greece) and Hammerfest (Norway) (Fig. 1). Of these 33 (and one supplementary site) had Urtica dioica present and a herbarium voucher specimen was collected from each of these sites (and in addition one specimen of U. membranacea from Greece). Herbarium voucher specimens are deposited in the herbarium of the Natural History Museum, London (BM). Living specimens were also collected for cultivation in London (Queen Mary University of London), for future experimental work. One living specimen was collected from each site (two from site 27). Of the living specimens collected, 27 survived into cultivation and could be used for flow cytometry (see results). The living specimens were grown in London in a 'common garden' (rooftop plant growth facility at Queen Mary University of London, Lat. 51.5234. Long. -0.0423).

Flow cytometry and buffers
Ploidy level (diploid vs tetraploid) was assayed using flow cytometry (

Results
The flow cytometry results are given in Table 3. In all, 24 plants have flow cytometry results consistent with tetraploidy whereas five plants, 7-5 (Bulgaria), 11-4 (Romania), while 31-12, 32-11, 34-6 (southern Finland), have results consistent with diploidy. The identity of these diploid plants was checked and confirmed as U. dioica sens. lat. As diploidy is often associated with stinglessness, information on the presence of stinging hairs was collected after cultivation in a common garden (Queen Mary University of London, QMUL) for one year (Table 4). Information on flowering time in the common garden is also given. Flowering time shows an overall correlation with latitude (generally with late flowering plants coming from Finland and Norway, although there are some exceptions (Table 4). Voucher specimens of both the original specimens and plants after cultivation (in common garden conditions for one year) are deposited at the Natural History Museum, London (BM).   Urtica phenotype in common garden (London). Fl. (flowering) time refers to category of flowering performance in 2016; 1 = early flowering (flowering before 16 May); 2 = mid-June (flowering by 10 June); 3 = late June (21 June); 4 = early July (2 July); 5 = late or not flowering (not flowering by early July). Stinging hairs refers to the typical number of stinging hairs per leaf; 1 = <10; 2 = 10-50; 3 = 50-100; 4 = >100. Numbers are given for: adaxial surface (first number)/abaxial surface (second number). At the tetraploid level, some variation in the estimated genome sizes was detected, with the northern populations tending to have higher 2C-values compared with the more southerly ones (Table 3; Fig. 2). To confirm that this intraspecific variation was genuine rather a technical artefact, leaves from the two individuals showing the largest difference in 2Cvalue (i.e. 28-10, 37-6) were co-processed. This resulted in two distinct peaks in the flow histogram (Fig. 3), indicative of biologically real difference in C-values at the tetraploid level. Scatter plot of genome size (2C-value, pg) values (as given in Table 3), plotted against latitude (Table 3). Only tetraploids (4x) are shown; diploid samples (2x) are not plotted. Note the generally higher genome size of the high latitude samples (see Table 3).

Discussion
The results confirm that the tetraploid is the dominant cytotype in our sample of U. dioica but that diploid plants do occur relatively frequently (at least in SE Europe and S. Finland). A more extensive survey of cytotype variation in Romania and Bulgaria, as well as around the Baltic would be of interest. Ploidy level has been shown to correspond with morphological characters (Geltman 1984;Geltman 1986  Screen shot from the Partec CyFlow flow cytometer showing flow histogram obtained from analysing Urtica dioica accessions 28-10 (peak 1, 2C=2.36 pg) and 37-6 (peak 2, 2C=2.59 pg) showing two distinct peaks and hence demonstrating genuine intraspecific variation in genome size between these two tetraploid individuals (28-10 and 37-6: see Table 3). The graph shows the relative fluorescence (indicative of DNA amount) in thousands of cell nuclei. The machine also gives summary statistics for the peaks. Note the very low coefficient of variation (CV%) of 2.48% and 2.89%.
pubescence. Until the populations from which these plants come can be examined critically we tentatively assign all our samples to the variable U. dioica ssp. dioica.
The finding of diploids in SE Europe raises the possibility that the widespread tetraploid form of Urtica dioica subsp. dioica, which has also become a weed, may have originated there, and the diploids may have survived glacial episodes in S. European refugia. The origin of the diploids of S. Finland is as yet unknown, although a phylogeographic analysis might be informative here. Another interesting result is the discovery of intraspecific Cvalue variation, particularly the generally higher C-values in the far north. This may be indicative of a distinct population of nettles in the north, and again this would benefit from more detailed cytogeographic study.