Fertilization can affect vegetation dynamics and natural grassland diversity. This study evaluated the vegetation dynamics of a natural grassland 16 years after the initial fertilization, discussing the long-term effects of addition of triple superphosphate (TP) or Gafsa rock phosphate (RP) sources, as well as the effect of exotic species introduction on the inter seasonal dynamic of floristic composition. Phosphate (P) was applied in 1997, 1998, 2002, 2010, and 2012 at the quantities of 78.6, 39.3, 43.7, 43.7 and 43.7 kg·ha^-1, respectively, totaling 249 kg·ha^-1 P. Total herbage mass production (THM) with RP (13 485 kg·ha^-1) and TP applications (14 668 kg·ha^-1) was higher than in the Control (11 291 kg·ha^-1). There was a higher warm tussock perennial grasses C₄ contribution on herbage mass (HM) during the summer season (1 106 kg·ha^-1), whereas it was similar between treatments. In summer, the warm-season prostrate perennial grasses C₄ group contribution for HM was on average 48% higher when RP was used (1 590 kg·ha^-1) in relation to the other treatments. The HM contribution from the cool season annual grasses C₃ group (CAG) in the total HM, over spring 2012, winter and spring 2013 in TP treatment, was 17% higher than the other treatments. The changes in the seasonal botanical composition dynamics mainly by inducing modifications in the proportion of Paspalum notatum A. H. Liogier ex Flüggé on RP treatment and Paspalum urvillei Steud. and Lolium multiflorum Lam. on TP treatment. However, no significant effects were observed in species richness, which ranged from 19–24 species among growth seasons. In the same way, the Shannon Diversity Index and Pielou Equitability Index were not modified by historical P sources. These results indicate that phosphorus fertilization has lower effects on natural grasslands diversity and could be used as a tool with important implications for livestock.