Perennial plant species diversity and its socio-ecological determinants in selected agroforestry practices at eastern escarpments of Chercher Massive, Ethiopia

The majority of studies in agroforestry have primarily focused on boosting agricultural production, income generation, component interaction and less emphasized on the aspect of biodiversity. Several studies have called for more investigation to address the lack of understanding regarding the significance of agroforestry in conservation of plant biodiversity. This paper examined the role of agroforestry practices in conservation of perennial plant species diversity and the relationship that the perennial plant species diversity has with altitude, slope, and household wealth status. Data was collected from 216 sample plots of agroforestry practices using a multi-stage sampling system. The relationship between diversity indices and explanatory variables such as agroforestry practices, household wealth, altitude, and slope gradients were assessed using a generalized linear mixed model. The impact of these variables on plant species composition was evaluated using the Adonis2 function. Sorenson dissimilarity measures were used to determine the dissimilarities in species composition. A total of 135 perennial plant species with 58 families were identified, with natives accounting 68.15% of the total. The Fruit-tree-based agroforestry practices maintained significantly higher observed and expected number of species (p < 0.05). The values of the mean species richness, Shannon diversity, and Margalef indices per plot, were found to be 6.1 ± 0.65, 1.53 ± 0.32 and 6.37 ± .045 respectively. The dissimilarity in species composition between the two agroforestry practices was 0.53 while the Sorensen multiple-site dissimilarity measures among the kebeles was 0.72. The coffee-shade-tree based agroforestry practices deserved higher importance value index of species. The species richness, Shannon diversity and the importance value indices were significantly differed with respect to slope, altitude, agroforestry practices and household wealth status (p < 0.001). The species composition varied along altitudinal gradients, household wealth status and across the study kebeles (p = 0.005, 0.013, and 0.029, respectively). Finally, it is argued that agroforestry practices should be given more emphasis as a way of perennial plant species reservoirs in relation to the various socioeconomic and ecological factors in order to counterbalance the local threat of perennial plant species from their natural habitats.