• More than half of trees species in Kihamba system are non-native • Nativeness of the tree species influence structural diversity of the tree layer • Native tree species in Kihamba are at risk of extinction Agroforestry systems in mountainous landscapes deliver essential ecosystem services; however, changes in the overstory trees may threaten their structural integrity and resilience. We assessed the introduction of non-native tree species on the Kihamba agroforestry system of Mt. Kilimanjaro to evaluate their role in structural diversity and biodiversity support. Data were collected from 30 Kihamba farm plots across elevational belts in three transects representing different socio-economic contexts. Analyses included species abundance, diversity, tree-layer heterogeneity, and regeneration patterns, contrasting the main native and non-native taxa. We found that more than half of the tree species in the Kihamba system were non-native, with higher concentrations in upper elevational belt. At plot level, Persea americana and Grivellia robusta (non-natives) dominated, followed by Albizia schimperiana (native), indicating a mixed assemblage of indigenous and cultivated species. Species richness and evenness showed no significant variation between elevational belts or among transects, but the lower elevational belt exhibited greater crown cover and structural diversity. Structural diversity was driven by native canopy dominance and total basal area rather than by species richness, indicating that ecosystem structure depends more on which trees are present than on how many species occur. Native species also exhibited higher active regeneration (40.7%) than non-native species (27.8%). However, recruitment credit was higher for non-native species (72.2%) than for native species (51.9%) with 7.4% of natives showing extinction debt. Despite non-native prevalence, native species remain pivotal for structural diversity, highlighting the need to promote their regeneration to sustain ecological resilience in agroforestry systems.
Kalumbilo et al. (Sun,) studied this question.