details such as habitats surrounding the site are recorded. This survey is appropriate for
volunteers with little or no previous experience – including householders.
(2) Field survey monitoring. Volunteers are allocated a randomly selected grid reference along
a stretch of waterway or of a 1-km2 area. They select a route to walk along the waterway
or around the square, which includes stopping points spaced out appropriately. The route is
walked once during the day to record the habitats present. Two evening surveys walking the
route with a bat detector and powerful torch are made in July/August. Bats are recorded while
the observer is in transit and while he or she is standing still at the stopping points. This
method requires training in the use of a bat detector.
(3) Hibernation-site monitoring. Counts are made of all species encountered at a range of sites
selected by surveyors – typically caves, mines and cellars. Two survey visits are made, one in
January and one in February each year. Hibernating bats are identified and counted without
disturbance. A torch with a deep-red filter will help prevent disturbance in caves or buildings.
Photos can be taken and counted later to minimise disruption to colonies.
Warren and Witter (2002) showed that their monitoring programme in Wales could detect
a 5% population change over a 5-year period of monitoring. O’Donnell (2002) suggested
that 10 years might be a more conservative figure, given the high variability in night-to-night
counts for some species roosting in caves.
A variety of ultrasonic bat detectors has been used over the past 30 years to identify
free-flying bats (O’Farrell et al. 1999). Analyses of recorded echo-location calls with older
machines were often slow and typically restricted to few calls, but modern computing power
has allowed species identification from calls in real time. Use of the Anabat II detector
(http://www.titley.com.au/tanabat.htm) and its associated analysis system allows an immediate
examination, via a lap-top computer, of the time–frequency structure of calls as they are detected.
These calls can be stored on the computer for later examination. Many bats can be identified as
to species by qualitatively using certain structural characteristics of calls, primarily approximate
maximum and minimum frequencies. All bat calls are not equally useful for identification. To
identify calls precisely, it is important to use a continuous sequence of calls from an individual
356 Mammals
in normal flight rather than single isolated calls. Counts of free-flying bats with bat detectors are
subject to high variability, so with every monitoring programmme it is necessary to reduce this