Biodiversity loss due to human activities is likely to increase due to projected increase of human populations. Humans have transformed forest, grasslands and savannas to produce food, fiber and timber (DeFries et al. 2004, Ramankutty et al., 2008). Models of land conversion predict a reduction species richness by 40.3% globally (Newbold et al. 2015). The projected decrease in species richness might be underestimating the loss of unique or similar species because those measures assume all species are equal. Phylogenetic diversity not only accounts for the number of species present but also for their evolutionary relationships. Phylogenetic diversity is a biodiversity index that can reflect the diversity of features or characteristics represented in a community (Faith 1992), thus a community that lost phylogenetic diversity would lose functional diversity. Functional diversity refers to the value and range of species traits in an ecosystem, such as leaf size, nitrogen content, and canopy heights (Díaz and Cabido 2001). Even in areas where species richness remains the same, it has been suggested that agricultural history has altered community composition and reduced phylogenetic diversity (Turley and Brudvig 2016). We examined the effects of human activities on phylogenetic diversity and structure in the longleaf pine savannas. We quantified measures of phylogenetic diversity for plant communities described by vegetation surveys conducted in the uplands of the Red Hills Region of Florida and Georgia during 2017 and 2018 (Dixon et al. 2021) as part of study investigating ecosystem services in the area and their relationship with species richness. The areas surveyed comprise 10 different land use types including annually tilled row-crop fields which were all tilled within the year, five different periods of time since abandonment (5-15 years, 15-30 years, 30-50 years, 50-75 years, and 75-100 years, which we will refer to as restored savanna), and reference sites (never-plowed, frequently burned pine savanna). There were 2 to 4 replicate 100 m2 plots per land use type at each property for a total of  101 plots.  The plots contained more than 386 plant species. Phylogenetic diversity doubled within five years of fire management. Continuous fire restoration helps recover phylogenetic structure after 15 years (mean nearest neighbor taxon distance, mean pairwise taxon distance). Communities affected by human activities species tend to be less closely related than in never plowed communities. Species that did not return after anthropogenic disturbance were not a random subset of species. One explanation for this clustering is that abiotic filtering influences recovery after disturbance.