Demographic responses to habitat fragmentation: Experimental tests at the landscape and patch scale

We tested the ecological consequences of habitat fragmentation by comparing the density, population growth rate, survivorship, and recruitment of Microtus pennsylvanicus populations within a 20-ha fragmented landscape with those of populations in a 20-ha unfragmented landscape. We also tested for fragment-size effects by comparing the same measures of demographic performance across three fragment sizes (0.06, 0.25, and 1.0 ha). During 17 censuses between June 1993 and October 1994, we recorded 10 020 captures of 3946 individuals and found strong landscape differences but weak fragment-size effects. Although fragmentation reduced the habitable area by 72%, density and adult recruitment were significantly higher on the fragmented landscape relative to the control. With the exception of adult recruitment (higher on small relative to medium and large patches), no significant demographic differences existed among patches of different size. Low rates of between-population movement and differential juvenile growth rates suggested that higher recruitment rates on the fragmented landscape likely resulted from enhanced local reproduction rather than from immigration. Thus, despite the fact that populations in the fragmented landscape experienced severe habitat loss, some individuals on fragments accrued important reproductive advantages (possibly as a result of diminished social costs or enhanced food resources). That population and individual responses to fragmentation could differ so dramatically provides a novel result that illustrates the importance of using hierarchical field designs in tests of population responses to large-scale habitat alteration. We conclude that controlled, large-scale field tests can serve as an important intermediary between the inherent abstraction of simulation modeling and what is observed in the real world.