1. Globally the prevalence and impact of invasive non-native plant species is increasing rapidly. Experimentally-based research aimed at supporting management is limited in its ability to keep up with this pace, partly because of the importance of understanding historical abiotic and biotic conditions. Contrastingly, landholders are in unique positions to witness species turnover in grasslands, adapt management practices in response, and learn from successes and failures. 2. This local knowledge could be crucial for identifying feasible solutions to land degradation, and ecological restoration, but local knowledge is rarely explicitly embedded in ecological research. 3. In this study, we use a sequential exploratory strategy where we first interview (semi-directive approach) 15 landholders within the Bega region of New South Wales, Australia concerning the changing ecological characteristics of both extensively and intensively managed grassy woodlands and perceived impacts following arrival of the invasive exotic introduced species, African lovegrass, Eragrostis curvula. 4. Based on the results of these interviews, we then conducted a field study at 57 sites where we tested seven landholder-generated hypotheses. 5. The field study validated many of the landholder management perceptions including: African lovegrass was negatively correlated with species richness, canopy cover and dominant grasses like Themeda trianda. Mechanical slashing increased exotic African lovegrass abundance. The prevalence of African lovegrass in the soil seed bank was positively correlated with its abundance aboveground. Study observations that contradicted landholder perceptions included: African lovegrass was not more palatable nor did its’ abundance decline in response to increasing soil fertility. Spot spraying with herbicides was effective at controlling abundance, despite its reputation as ineffective. Landholder observations also highlighted key hypotheses concerning modes of spread that require long-term studies including the roles of drought and overgrazing. 6. Synthesis and applications. Overall, we found local knowledge coupled with scientific methods can act in tandem as a highly effective approach for developing management recommendations. This approach identifies local perceptions that are not substantiated by scientific data to halt potentially harmful practices, and observations that are insightful predictions about the dynamics and impacts of non-native species that need long-term experiments to corroborate scientifically.