Context. Invasions by non-native plants can alter the abundance and distribution of resources that can affect habitat quality for native animals.

Aims. We sought to understand the demographic consequences of a plant invasion on a functionally and numerically important rodent in a grassland ecosystem. Specifically, we evaluated how abundance, survival, reproductive activity and population structure of Arizona cotton rats (Sigmodon arizonae) varied across a gradient of invasion by Eragrostis lehmanniana (Lehmann lovegrass), a bunchgrass native to Africa that has invaded grasslands in North America.

Methods. Over a four-year period, we used capture–recapture methods to survey small mammals on 54 1-ha plots between 10 and 13 times. We used vegetation data collected each autumn to quantify biomass of non-native grass, total biomass and vegetation heterogeneity to characterise vegetation structure on each plot.

Key results. We captured 1344 individual cotton rats during 106 560 trap-nights across all sampling periods. In areas dominated by non-native grass, abundance of cotton rats increased 7- to 10-fold and survival increased by 117% relative to areas dominated by native grasses. In contrast, reproductive activity of adults decreased by 62% for females and 28% for males, and the proportion of adults in the population decreased by 20% in these same areas.

Conclusions. Demography of Arizona cotton rats differed markedly in areas invaded by a non-native plant relative to native grasslands, supporting the long-held idea that life histories can reflect local environmental conditions. Because distributions of many non-native plants are predicted to increase in response to future changes in natural and anthropogenic drivers, the potential breadth of these complex effects on communities of native animal is immense.

Implications. The complex variation in demographic responses across the invasion gradient suggests that it may be necessary to evaluate a suite of vital rates to fully understand the consequences of plant invasions on animals. This is especially important for species of conservation concern because single demographic parameters, which are used frequently as targets to gauge the success of conservation and management activities, could be misleading.