To test the hypothesis that voles reduce the amount of time spent burr
owing when risk of predation is low, we excluded predators from four p
enned populations of voles and allowed access by predators to four oth
er penned populations. Voles responded as expected; pens without preda
tors had fewer entrances to burrows and fewer simple burrows (short, b
lind tunnels and escape burrows). Spatial distributions of entrances t
o burrows remained random or clumped in both treatments. Multiple-nest
burrows were larger and more complex than single-nest burrows or esca
pe burrows, but the structure within each type of burrow did not diffe
r between treatments. Although voles altered the types of burrows they
constructed in response to risk of predation, some other factor appar
ently constrained them from changing the structure of their burrow sys
tems.