Geoscience Reference
In-Depth Information
Wandering−quarter sampling.
Japanese black pine saplings
1.0
0.8
0.6
Y
0.4
0.2
0.0
0.0
0.2
0.4
0.6
0.8
1.0
X
FIGURE 6.5
One possible placement of transects (arrows) and derived trajectories (determined by the direc-
tion of the arrows) inside quarters (delimited by the dashed lines) for the measurement of dis-
tances between Japanese black pine saplings, sampled by the wandering-quarter method. The
cross × on each transect line indicates a random starting point for the search of the nearest sap-
ling inside a quarter. The estimated density [using Equation (6.6)] is 63.55 saplings per unit area.
procedure involving the measurement of angles between the lines of sight
from sampling points to their nearest two neighboring objects and the cor-
responding estimation of the mean angle as an index for testing spatial ran-
domness. Furthermore, Assunção and Reis (2000) compared the performance
of distance-based tests of randomness based on
T
2
distances and angle-based
tests and found that the Hines and O'Hara Hines procedure was the best
test in terms of power. Nevertheless, Assunção's angle-based procedure has
been recommended by Trifković and Yamamoto (2008) for indexing spatial
patterns of trees in forest stands, but they warn that this indexing can be
improved when the mean of angles is accompanied by a description of the
frequency distribution of angular measurements.
T
T-square sampling and the wandering-quarter method are relatively
robust to the assumption of a random spatial pattern of items. However,
there has recently been some interest to apply design-based approaches
for density estimation in plotless sampling, without assuming a particular
