POY analysis

Results of POY analysis are especially interesting for comparison because they are produced without the subjective stage of manual aligning. The details of POY analysis are described in Gardiner et al. (2005). 

Trees were rooted on Hookeria (Hookeriales) and with the other representatives of the Hookeriales included in the analyses (Lopidium in Nona and POY analyses, and Distichophyllum in POY analysis) formed a grade basal to the other taxa included, all members of Hypnales according to Goffinet and Bnck (2004). 

As the POY analysis was based on a slightly different dataset, its clades in many cases are not identical in species composition, although these differences are generally rather small. Thus, we apply the same abbreviation for POY topology, marking clades with slightly different species composition as, for example, 03, 04 instead of 03, 04. 

POY analysis was based on a dataset with 9 species excluded from the subsequent analyses thus some differences were expected. Nevertheless, the main topology of the tree (01, 02, Ml, M2) was the same, although 02 was expanded by a clade (sister to Ml + M2), that included Callicladium... 

Analysis of x ray dii action results (Figure 8) showed that there is only p-fdiase in the samples fl7 FDY, G4 FDY, f72 FDY 70 C and f72 FDY 160 C. A mixture of a-and P-phase was found for the sample Snowflake FDY and pure a- [jliase for f72 POY and Snowflake POY. The same trends were also confirmed by Raman >ectroscopy. This is consistent with etq>ected results from literature on phase tranrfmmations of PVDF. 

Once the main clades within the Brachytheciaceae had become more or less stabilized, our main goal was to elucidate the position of taxa not included or not well represented in the previous analyses, and also to find additional arguments for generic and subfamilial placement of some particular taxa. Three methods are involved here (1) evaluation of the molecular synapomorphies in POY alignment, i.e., characteristic indel events and substitutions (2) analysis of the secondary structure of the tmL intron and (3) phylogenetic analysis of a reduced dataset from previous analyses, with some additional taxa. 

All fonr trees inclnde a basal grade (01-02) and two main clades (Ml and M2). The composition of the basal part of the basal grade (01) and the general composition of the two main clades were fonnd to be stable in MB 135, MB 144 and Nona analyses. In the distal part of the basal grade (02) one species from 02 moved to Ml and one from Ml to 02 as a result of the inclusion of additional taxa in the analysis. In the POY tree, which, however, was based on a somewhat different dataset (as discnssed below), most of the differences compared with Nona and MB analyses were in the topology of the distal part of the basal grade, 02. 

---