Tropical botanical gardens have played an important role in the distribution, naturalisation and spread of non-native plants worldwide. Appropriate guidance relating to risk assessments of established botanical garden collections is often scarce. This paper uses the Amani Botanical Garden (ABG), Tanzania, as a case study to highlight appropriate methods to assess the risks posed by existing and future collections in tropical botanical gardens. Key considerations included Weld assessments of species status using accepted deWnitions of naturalisation, spread and invasion, distinguishing between intentionally and unintentionally introduced species, identifying taxonomic patterns in invasion status, assessing patterns in habitats colonised, and determining how knowledge of invasion elsewhere might be useful in forecasting risk. Out of the 214 alien plant species surviving from the original plantings in the early 20th century, 35 had only regenerated, 38 had locally naturalised while 16 had spread widely in the botanical garden. A further 16 species with unclear introduction records in the garden were also found to be naturalised. A greater proportion of introduced species were potentially invasive than might be expected from previous analyses of global Xoras. Overall, just over half of all naturalised and spreading species were also observed in forest fragments and edges. The proportion of species that had been recorded elsewhere as naturalised/invasive was signiWcantly related to their status in ABG, 1 C with 94% of spreading species and 79% of naturalising species being recorded as naturalised or invasive elsewhere, compared to 57% of species that were only regenerating and 49% of species only surviving. Recommendations for further risk assessments of botanical garden collections are discussed.

There is insufficient information regarding the factors affecting the environmental impacts of alien species. In particular, little is known about whether there is any relationship between the invasiveness (establishment and spread) of an introduced species and its per capita impact. We experimentally assessed the relationship between the extent of spread of up to 29 alien plant species and their impact on recruitment of native tree species in Amani Botanical Garden, Tanzania. We also studied the effects of allelochemicals of selected alien on native plant species to assess potential mechanisms of impact. We found no relationship between the extent of spread of an alien tree species and their impact on seed germination, seedling survival, and seedling communities of native trees in their understory, and no indication that allelochemicals consistently explain their effects on recruitment of the studied species. These results suggest that extent of spread cannot be used as a proxy for impact. Hence, managers should continue assessing both the spread and the impact of alien species when prioritizing alien species for management.