Soil macroaggregate stability under different fallow management systems and cropping intensities in southwestern Nigeria

Evaluation of soil aggregate stability under managed fallow systems is very relevant in the assessment of their agricultural sustainability. It is als o important to identify the parameters that are most responsive to the effe cts of these fallow systems on aggregate stability. This study was carried out in southwestern Nigeria to quantify the effects of various fallow manag ement systems on the macroaggregate stability of surface soil (0-15-cm dept h) using a long-term fallow management trial established in 1989 on an Alfi sol toposequence. Data were collected between 1994 and 1995. Three fallow s ystems (bush fallow, Pueraria phaseoloides, and Leucaena leucocephala) and three fallow periods (1-3 years after 1-year cropping) were evaluated. Crop ping consisted of maize + cassava intercropping. A split-plot design with f ellow systems as main plot and cropping intensities as subplot in four repl ications was used in the trial. Soil aggregate samples were wet-sieved to e valuate their mean-weight diameter (MWD) and fractal dimension (D). The mea ns of MWD for the fallow systems and cropping intensities ranged from 2.4 t o 6.4 mm. The mean D values ranged from 2.29 to 2.72 while the mean interce pts (log k) of the regression ranged from 2.01 to 2.28. Low D values (cohes ive and stable aggregates) were associated with fallowing whereas high D va lues (fragmented aggregates) were associated with cultivation. Fallowing en hanced soil aggregate stability. The Pueraria system also enhanced soil agg regate stability more than the bush fallow and Leucaena systems when contin uous cropping was practised. The subplots, which were fallowed for 2 or 3 y ears after 1 year of cropping, were usually similar in aggregate stability to the secondary forest soil. Fractal analysis showed further that soil agg regates in the dry season and on the upper slope of the toposequence were m ore cohesive than in the wet season and on the lower slope. Observed D and log k values were also significantly influenced by laboratory method. (C) 1 999 Elsevier Science B.V. All rights reserved.