| Abstrakt: |
Trees on agricultural fields can have a positive or negative impact on weed seedbank (WSB) due to diverse environmental and soil characteristics. Therefore, soil samples were drawn in six cropping systems [two agroforest systems (AFS): guava, mango; three horticulture systems (HCS): guava, mango, Indian gooseberry; and annual crop system (ACS)] at two landscape positions (lowland and upland) and two soil depths (0-15 cm and 15-30 cm) using factorial randomised block design each replicated three times. Results showed that guava-AFS had the highest WSB of different categories in general and individual weed species in particular, except for Eragrostis pilosa and Dactyloctenium aegyptium. Simultaneously, guava-AFS also showed the maximum Shannon-Weaver, species richness and Simpson index and was low in Whittaker statistics (ßW). The species evenness varied non-significantly with the cropping systems. Similarly, the landscape position had no discernible effect on any weed diversity indices; however lowland landscape position was dominated by Cyperus spp. and E. pilosa, while the upland by Phyllanthus niruri. Furthermore, with the exception of ßW, the WSB and diversity indices were found to be higher on the topsoil (0-15 cm). Our study establishes that the AFS system in the semi-arid sub-tropics has a more diverse WSB indicating a healthy system, as opposed to HCS, which has a dominance of certain weed species, opening the door for more severe infestation of invasive weed species. [ABSTRACT FROM AUTHOR] |
