An initial explanation was that these trees might represent forest ingrowth, which could be verified if the tree diameters were generally small. However, quick analysis of these trees' dbh measurements show no noticeable difference between their diameters and the diameters from the rest of the population. Graphs 2.1 and 2.2 show the distribution of these trees' dbh values.
Previously, these trees were treated like the trees with missing data. (The Fortran program interpreted missing data as zeros and thus was unable to distinguish between these two types of trees.) The same spacing problem described in the previous section applied to these trees. Where 36% of the trees1 [with correct position data] were omitted from the final data set, none of the trees with angle and radius of zero were omitted.
These trees could either be treated like the trees with missing data (the mortality trees described above) or they could be treated in a different (presently unspecified) manner.