Geoscience Reference
In-Depth Information
Figure 7.1
Preliminary results from the River Habitat Survey baselines: (a) the extent (percentage of river length) of
channel resectioning in 2007-2008 baseline, using ordinary kriging to interpolate between the 4884 RHS sites; (b) the
change in distribution, based on occurrence in 10 km Ordnance Survey grid squares, of Himalayan balsam (
Impatiens
glandulifera
) between 1995-1996 and 2007-2008.
©
Crown copyright, Ordnance Survey. All rights reserved.
climatic correlates of salmonid densities (Clews
et al
., 2010); (iv) national studies of river channel
form (Emery
et al
., 2004; Harvey
et al
., 2008);
(v) regional studies of nutrient concentrations and
fluxes (Jarvie
et al
., 2003); and (vi) appraising river
management interventions (Clews and Ormerod,
2010).
There are two main challenges to broadening
the use of river monitoring data. First, the
data constrain the questions that can be asked,
based on the subject matter, sampling method,
and geographical and temporal coverage. In
effect, this is an unavoidable side effect of
post
hoc
analysis because it reverses the scientific
method of first defining the question and then
collecting appropriate data. Second, most of the
study parameters will be fixed
aprior
, with
a different purpose in mind. This encompasses
fundamental issues such as the taxa sampled, and
the level of identification (e.g. family versus species
identification of macroinvertebrates),
locations and frequency, and the suite of variables
measured (Vaughan and Ormerod, 2010). A
common consequence is that while monitoring
data may include aspects relevant to the
post hoc
study, they are not optimal in answering the
desired question. For example, in relating biology
to water chemistry, episodic high concentrations of
certain determinands may be crucial (Kowalik
et al
.,
2007), but routine monthly chemical sampling
is unlikely to be frequent enough to distinguish
such events. Similarly, general characterization,
using survey methods such as RHS, may not be
optimal for recording the habitat preferences of
particular taxa or looking at specific aspects of
channel morphology (Hastie
et al
., 2003). In these
cases, analysis of monitoring data may provide
some useful supporting or contextual evidence,
perhaps confirming known theory (Hastie
et al
.,
2003; Vaughan
et al
., 2007), but it cannot address
the specific issues in the same way as a properly
designed investigation.
sampling
