Plant diversity and soil seed bank composition in scattered tree-dominated farmlands, Southern Ethiopia

Abstract

In parkland agroforestry, assessing the soil seed bank (SSB) provides valuable insights into the potential for natural regeneration and recovery of plant communities following disturbance. This study evaluated the composition, spatial distribution, and regenerative potential of SSB beneath and outside the canopies of Acacia tortilis , Balanites aegyptiaca , and Cordia africana in Southern Ethiopia. Standing understory vegetation was evaluated using 144 quadrats (1 m × 1 m), while soil samples were extracted from 0 to 20 cm and 20 to 40 cm depths for controlled seedling emergence trials. Analysis of Variance (ANOVA) revealed that seed attributes were significantly governed by a three-way interaction between tree species, canopy zone, and soil depth. Propagule pools were intensely canopy-centric; mean seed densities beneath over-story covers (28.13 seeds m −2 ) more than doubled those of open outside matrices (12.65 seeds m −2 ), with A. tortilis driving the highest accumulations. Vertically, severe downward attenuation occurred, with topsoil (0–20 cm) maintaining significantly higher mean densities (32.53 seeds m −2 ) than subsoils (8.33 seeds m −2 ). Out of 28 germinated species across 13 families, the SSB was overwhelmingly dominated by opportunistic herbs (46.43%) and grasses (39.29%), whereas native woody species were severely suppressed (14.29%). Furthermore, Jaccard’s index revealed a weak compositional match between the subterranean core and standing vegetation. The scarcity of woody propagules suggests that the soil seed bank alone may be insufficient for spontaneous woodland recovery in managed farmlands, particularly given that this study was conducted at a single location and evaluated only germinable seeds. Consequently, complementary active restoration strategies, such as enrichment planting and direct seeding, should be considered to support regional biodiversity and long-term agroforestry productivity.

The authors' abstract, as published at the source. Agroforestry Systems, 2026 · DOI ↗