In Northern Nigeria, the savannah parklands are mixed dryland agroecosystems having landscape productivity and deforestation challenge, induced by the changing climate and affecting sustainable livelihoods. The productivity of parkland is attributed to different factors such as erratic precipitation, drought and plant mortality. Here, the study predicts the impact of climate change on tree regeneration on parklands of three savannah agroecological zones (AEZs), namely the Sudan savannah (SS), Northern Guinea savannah (NGS) and Southern Guinea savannah (SGS) AEZs across a north-south transect using a multidisciplinary approach. The approach involves the employment of Random Forest model framework to predict the future shift in AEZs from the current climate using bioclimatic variables and climatic scenarios. Three random predicted locations within each studied AEZ identified as drought-threatened areas were selected for tree preference and identification study by analysing local farmer ethnobotanical knowledge of their favourite species, before tree identification and estimation of diversity and abundance of the current parkland species on farms in the respective local communities. The research further predicted the future distribution of selected important species using Maximum Entropy and assess the efficacy of simple propagation methods (cuttings and marcotting) on respondents’ farms, after an extreme drought survival test in a greenhouse. The results showed that agroecological zones are getting warmer and drier southwards, replacing the current NGS and SGS with SS between 2050 and 2070, by up to 65%. Parkia biglobosa, Mangifera indica and Vitellaria paradoxa are the top-ranked species by 90% of the 92 respondents in the study and are the most abundant species on farmers’ field with relative species dominance ranging from 19.61-42.64%. The strong relationship between species preference and abundance increases northwards. On the future distribution and improvement of parkland species to survive under stress conditions. Maximum entropy model predicted a north-eastern shift for future spatial species distribution of Parkia biglobosa and Vitellaria paradoxa under most extreme climate scenario. For tree species at jvenile stage to survive under 6 months drought regime, the research indicated that 9months old Parkia biglobosa seedling from SGS resisted more than other species at different ages from different AEZ species. On local adaptation strategy to improve parklands survival, there were up to 40% and 38% success rate of vegetative propagation of Anarcadium occidentale using cuttings and marcotting, respectively.