Spatial distribution and population dynamics of Striga hermonthica seeds in naturally infested farm soils

Abstract

Densities and spatial distribution in soil of seeds of Striga hermonthica were analysed for four naturally infested farm fields in Western Kenya. A revised method for extraction of Striga seeds from soil was used, combining centrifugation with existing techniques based on flotation. Tests showed that 85% of Striga seeds were retrieved from soil samples. In all fields the majority of seeds were found in the plough layer (0 – 20 cm). New seeds entering the soil from the surface after seed shedding created a strong gradient with depth. Downward penetration from the soil surface was larger in sandy soil than in clay soil. In tilled soils no significant vertical density gradient was found within the plough layer. At a fine scale (0.2 m) seed densities showed little horizontal variation, but significant differences in seed densities in the horizontal plane were found at larger scale distances (several m) between locations in all fields. At 125 days after sowing the estimated average number of seeds produced per emerged Striga shoot was 4,827, excluding an approximately similar amount of seeds present in maturing capsules. The estimated average number of seeds produced per mature Striga seed capsule was 1188. Large seasonal fluctuations in the Striga seedbank were measured. An average net increase of 88,825 Striga seeds m-2 (equivalent to 340%) was calculated from seedbank analyses in 16 sorghum plots. The level of Striga infestation in one field had decreased by 62% from 34,250 seeds m-2 to 13,125 seeds m-2 after keeping it fallow for a year. A sharp decline in Striga seed density was found in samples taken at increasing distances from highly infested fields, irrespective of wind direction or slope, suggesting very limited dispersal of Striga seeds by wind or water. Parasite emergence was non-linearly related to initial Striga seed densities in the soil, but this relationship was only observable at the scale of individual plant holes. Seed production was also non-linearly related to numbers of emerged parasites, when measured at plot scale (25 m2), but not at the scale of individual plant holes. In the fields we studied, seed densities below levels of 13,000 Striga seeds m-2 could be considered to suppress the number of emerging parasites. However, if two or three emerged Striga plants per m-2 were left to seed, enough seeds would be produced to keep the seedbank in balance.