Problems in elm and rose Identifications
Fig. 4. Separation of Elms using leaf width/length ratios and no. of veins per leaf
Discussion I
The statistical comparison of elm leaves, taken from short shoots, confirms the impressions gained
from a general inspection of the trees. Those on the north-eastern and north-western boundaries of
the long-gone Beckling Wood are the most variable, and therefore probably belong to more than one
genetic clone. Elm in the study hedge shows far less variability, and may indicate that the trees here
arose from a single clone.
Elms of Willow Wood and north-east Witch Wood are relatively close, both in physical distance and
characteristics, and we assume that they are closely related. However, those from the southern
boundary of Witch Wood (some 600 metres away) have a mean value lying outside three standard
deviations of the north-eastern group from the same wood, and appear to be the most different from
al! the others.
Essex Naturalist (New Series) 18 (2001)
151