Non-native invasive species are altering ecosystems in undesirable ways, often leading to biotic homogenization and rapid reduction of evolutionary potential. However, lack of money and time hampers attempts to monitor the outcome of restoration efforts. Hence, it is useful to determine whether relatively limited sampling can provide valid inferences about biological responses to pattern-based and process-based variables that are affected by restoration actions. In the Mojave Desert, invasion of salt-cedar (Tamarix ramosissima) has altered vegetational communities and some measures of faunal diversity. We tested whether six vegetation-based predictor variables affected species richness of butterflies in the Muddy River drainage (Nevada, USA). We also explored whether similar conclusions about relationships between vegetation and butterflies could have been obtained by using data from a subset of the 85 locations included in the study. We found that the effect of non-native plants on species richness of butterflies was negligible. Availability of nectar had the greatest independent explanatory power on species richness of butterflies, followed by species richness of plants. In comparison with the full data set, subsamples including 10, 25 and 50% of sites yielded similar conclusions. Our results suggest that relatively limited data sets may allow us to draw reliable inferences for adaptive management in the context of ecological restoration and rehabilitation.