Although previous laboratory studies have commonly determined sprintin
g speeds of lizards on horizontal surfaces, the speeds and slopes used
during the escapes of lizards in natural habitats with variable incli
nes are virtually unstudied. To quantify performance and the use of in
clined surfaces during escape, we took advantage of the footprints lef
t in soft substrate and the simple surface topography of the natural d
une habitat of the Mojave fringe-toed lizard, U. scoparia. The lengths
of 52 escape paths ranged from 1.7-34 m, and the inclines of I-m port
ions of escape paths ranged from -25 degrees to 28 degrees, which effe
ctively encompassed the entire range of inclines in the habitat. The c
umulative frequencies (N=550) of inclines along l-m intervals of the e
scape paths were not a simple random sample of the habitat. Less than
10% of the cumulative distance travelled during escape was bipedal, an
d the escape paths were relatively straight. Trajectories of the first
metre of escapes were oriented significantly away from the presumed t
hreat (observer) but were random with respect to the orientation of bo
th the nearest cover and steepest incline. Eleven per cent of the cumu
lative number (N=1382) of strides measured were within 90% of the maxi
mum stride length within each path. Multiple regressions revealed that
stride lengths (and hence speed) during escapes in the field were max
imized on level surfaces with no turning. For lizards tested on a leve
l racetrack in the laboratory, maximum speeds averaged 2.8 mis (range=
2.1-3.9) and approximated 75% of the maximum performance attained in t
he field. (C) 1998 The Association for the Study of Animal Behaviour.