Geoscience Reference
In-Depth Information
method was the least biased, whichever allocation procedure of points, ran-
dom or semisystematic, is used. Aerts et al
.
(2006) applied
T
T-square sampling
to study the effects of pioneer shrubs on the recruitment of the tree
Olea euro-
paea
ssp.
cuspidata
in an Ethiopian savanna. Specifically, they recorded nurse
plants of seedlings using
T
T-square plotless sampling.
The
T
T-square method has been modified to allow the estimation of human
population sizes, particularly for estimating the size of displaced human
populations in emergency situations (Brown et al., 2001; Noji, 2005; Grais
et al., 2006; Bostoen et al
.
, 2007). As an example, Henderson (2009) consid-
ered occupied houses as the objects of interest, and instead of the density,
the number of occupants in houses was counted.
F i in a l ly,
applications and
alternative procedures for testing spatial patterns of objects sampled by the
T
T-square method were presented by Ludwig and Reynolds (1988), Diggle
(2003), Bostoen et al. (2007, Appendix 1), and ErfaniFard et al
.
(2008).
6.5 The Wandering-Quarter Method
The wandering-quarter method is an adaptation of the plant-centered quar-
ter method, not requiring any assumption regarding the spatial pattern of
the population studied. In fact, it provides a means of detecting regular,
random, or contagious distributions. In Catana's (1963) original proposal,
the data are collected from four transects positioned in the study area and
grouped in two sets of two parallel transects, each set being perpendicular
to the other. On each transect, the wandering-quarter method starts with
an initial point-to-object distance so that the nearest individual within a 90°
angle of inclusion around a compass line in the direction of the transect is
selected as the starting point for measurements. This first point-to-object
distance is followed by a chain of object-to-object distances so that another
90° angle of inclusion is built with the first object as a vertex and the com-
pass line as a bisector. The distance to the nearest individual within this
90° angle is recorded; the procedure continues until one of the boundaries
of the sampling area is reached or when a preestablished number of objects
have been found (typically 25). See the work of Catana (1963) and Tongway
and Hindley (2004) for more details. Figure 6.4 exemplifies the wandering-
quarter sampling process for a single transect.
Let
dd d
,
,
,
12
be the
N
individual wandering distances between items,
measured for the four transects combined, and let
…
N
∑
N
d
=
d N
/
i
i
=
1
