The efficiency of a vacuum device for estimating soil-surface seed abundance on lowland farms

Accurate sampling methods for estimating soil-surface seed abundance in agro-ecosystems are vital for understanding arable plant population dynamics, as well as quantifying seed availability as a food resource for farmland birds, mammals and arthropods. We tested the sampling efficiency of a portable, petroleum spirit powered leaf blower-vacuum in recovering four different seed size classes on the soil-surface of seven lowland farm habitats. Vacuum sampler efficiency was dependent on seed size, habitat type and exposure of seeds to environmental conditions. Larger seeds were more readily recovered and overall seed recovery was consistently higher on habitats with low vegetation cover, such as newly-sown cereal fields. We discuss the limitations of the device, but commend its use with correction factors to estimate soil-surface seed abundance in agricultural habitats. This study highlights the potential biases when estimating soil-surface seed abundance, the importance of correcting such biases and the need for better sampling techniques for weed science.