Climate Change: Defining Reference Conditions and Restoring Freshwater Ecosystems - Climate Change Impacts on Freshwater Ecosystems - page 216

The latter is particularly important as many European countries move towards
restoring the ecological quality of inland waters as mandated by the Water
Framework Directive.
Detecting recovery with changing baseline conditions
A fundamental part of ecological assessment in general, and restoration in
particular, is to isolate change induced by degradation/mitigation measures from
natural spatial and temporal variability. Statistical power to detect change can be
increased by reducing natural variability as much as possible (e.g. by stratifying
sampling effort) and increasing the signal or effect size by selecting variables with
strong responses (e.g. Johnson 1998; Sandin & Johnson 2000b). However,
adequate controls and replication are equally crucial for determining the
effectiveness of restoration.
Although at first glance, detecting human-generated disturbance, and assessing
the efficacy of restoration endeavours may seem to be two sides of the same coin,
there is one important difference. In determining ecological degradation, the null
hypothesis is that there is no difference between the putative perturbed site and
target, whereas for restoration, the criterion for success is to find no difference
between the putative restored and the target state (e.g. Downes et al . 2002). In
essence, this means that we have to test our hypothesis and not the null hypothesis.
This quandary can be circumvented if we hypothesize that during the extent of
the study, the putative restored site will differ from the initial or starting condition.
Ideally, such a statistical design would include three types of sites: (i) restored
sites; (ii) target or control sites; and (iii) sites similarly impaired as those restored
but not restored. A number of factors, such as study design, the confounding
effects of natural, spatial and temporal variability and a poor understanding of
the mechanisms driving recovery affect our ability to determine the success of
restoration.
Changing baselines confound interpretation
of recovery from acidification
International agreements and actions to protect and restore natural resources
threatened by acidification have resulted in marked reductions in the emission
and deposition of acidifying compounds (Stoddard et al . 1999) and concomitant
increases in surface-water pH (Stoddard et al . 1999; Lynch et al . 2000; Skjelkvåle
et al . 2000, 2003). However, despite putative recovery of surface water chemistry,
records of biological recovery are scarce and results are equivocal (Skjelkvåle
et al . 2000; Alewell et al . 2001; Stendera & Johnson 2008). Consequently,
acidification is still considered as one of the foremost problems affecting the
biodiversity of inland surface waters in Northern Europe (Johnson et al . 2003;
see also Chapter 7) and elsewhere (e.g. Kowalik et al . 2007; Burns et al . 2008).
Acidification often affects aquatic biota in a predictable way (Økland &
Økland 1986), even though the processes and mechanisms governing degradation
are not always clearly understood (Hildrew & Ormerod 1995; Strong & Robinson