Aim To determine the relative influence of medicinal use, height and geographical range on the chronological sequence of the first verifiable printed records of 1239 native species of flowering plants in the UK.
Methods We used Cox proportional hazards models to provide a direct estimate of the influence through time of explanatory variables on the hazard function.
Results In the period from 1538 to 1550, medicinal plants were 5–15 (95% confidence interval) times more likely to be discovered than non-medicinal plants. By 1600, 75% of medicinal plants had been discovered, and subsequently medicinal use had no significant influence on the probability of discovery. From 1538 to 1983, a 100-hectad increase in area resulted in a 4–6% increase in the probability of discovery. There was a small but significant decrease in the influence of area on the probability of discovery over the entire time period. In the same time period, a 10-fold increase in height resulted in a 3–35% increased probability of discovery.
Main conclusions Our results demonstrate that the first written records (1538–1600) of UK flowering plants were very strongly influenced by the perceived medicinal properties of the plants.