Description Usage Arguments Details Value Author(s) References See Also Examples
Computes effective strip width (ESW) for estimated detection functions from line transect data
1  ESW(obj, newdata)

obj 
An estimated detection function object. An estimated detection
function object has class 'dfunc', and is usually produced by a call to

newdata 
A data frame containing new values of
the covariates at which ESW's are sought. If NULL or missing and

Effective strip width (ESW) of a distance function is its
integral. That is, ESW is the area under the distance function from its
lefttruncation limit (obj$w.lo
) to its righttruncation limit
(obj$w.hi
).
Under perfect detection, area under the detection function is the entire
halfwidth of
the strip transect (from obj$w.lo
to obj$w.hi
).
Under perfect detection, density is the number sighted targets
divided by area surveyed, where area surveyed is
obj$w.hiobj$w.lo
times
total length of transects.
When detection is not perfect, less than the total halfwidth is effectively covered. Buckland et al. (1993) show that the denominator of the density estimator in this case involves total length of surveyed transects times area under the detection function (i.e., this integral). By analogy with the perfect detection case, this integral can be viewed as the transect halfwidth that observers effectively cover. In other words, a survey with imperfect detection and ESW equal to X effectively covers the same area as a study with perfect detection out to a distance of X.
The trapezoid rule is used to numerically integrate under the distance
function in obj
from obj$w.lo
to obj$w.hi
. Twohundred
trapezoids are used in the approximation to speed calculations. In some
rare cases, two hundred trapezoids may not be enough. In these cases, the
code for this function can be sink
ed to a file, inspected in a text
editor, modified to bump the number of trapezoids, and source
d back
in.
If newdata
is not missing and not NULL and
covariates are present in obj
, returned value is
a vector with length equal to the number of rows in newdata
.
If newdata
is missing or NULL and covariates are present
in obj
, returned value is a vector with length equal to
the number of detections in obj$dist
. In either of the
above cases, elements in the returned vector are
the effective strip widths for the corresponding set of
covariates.
If obj
does not contain covariates, newdata
is ignored and
a scalar equal to the (constant) effective strip width for all
detections is returned.
Trent McDonald, WEST Inc., tmcdonald@westinc.com
Buckland, S.T., Anderson, D.R., Burnham, K.P. and Laake, J.L. 1993. Distance Sampling: Estimating Abundance of Biological Populations. Chapman and Hall, London.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16  # Load example sparrow data (line transect survey type)
data(sparrowDetectionData)
# Fit halfnormal detection function
dfunc < dfuncEstim(formula=dist~1,
detectionData=sparrowDetectionData,
likelihood="halfnorm", w.hi=100, pointSurvey=FALSE)
# Compute effective strip width (ESW)
ESW(dfunc)
# ESW only applies to line transect surveys
# EDR is the point transect equivalent
# The effectiveDistance function tests whether the dfunc was
# fit to line or point data, and returns either ESW or EDR accordingly
effectiveDistance(dfunc)

Rdistance (version 2.1.3)
[1] 56.30084
[1] 56.30084
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.