Results were limited to texts available in English in peerreviewed
journals, where at least abstracts were available online.
Searches were cross-referenced to produce a single list for each
region. The abstracts of all papers were read and included in
the meta-analysis if they referred to the impacts of tourism, or
to services and/or activities related to the expansion of tourism
(e.g. desalination) where tourism was mentioned specifically as
the origin of the activity, or where the study area was known
to be a tourism area. In abstracts where this was unclear the full
text was read. Suitable articles were categorised under the following
topics: habitat loss, habitat damage, wildlife depletion,
wildlife disturbance and management of the environmental aspects
of tourism. Publications were also classified as quantitative
(contained empirical data on tourism and/or the potential impact),
modelling studies, or papers that listed purported impacts
but had no supporting data (e.g. reviews). The citation lists of
three representative papers from each category were also checked
for relevant studies that were not picked up by the database
searches.