Current methods for assaying the potential effects of rhizobacteria on plant growth are tedious and time consuming. A rapid, automated bioassay has been developed using a flatbed scanner and digital image analysis. Comparisons were made between manual measurements of root length and computed values based on an edge discrimination image analysis procedure. The assay consisted of bacterial cultures of filtrates added to 18 mL of 0.9% agar in square petri plates. Seeds were planted along a marked axis and incubated for 4 d at 15°C, and then the roots were measured. Total root length measurements for the two methods, manual and image analysis, were well correlated (r2 = 0.94), and both were equally able to identify bacteria inhibitory to plant growth. Root measurements by the scan method were five times faster than the manual method. Each sample required 53 s to read with the scan method vs. 250 s with the manual method. The scan method provides a quick and accurate method for screening inhibitory activity of rhizobacteria based on seedling root length and has potential in the screening of toxic substances of other origin.

Evapotranspiration (ET) of three perennial ryegrass (Lolium perenne L.) cultivars and one cultivar each of colonial bentgrass (Agrostis tenuis L.) and tall fescue (Festuca arundinacea L.) was measured in the field. Soil water depletion was measured with a neutron probe. Under minimal maintenance (i.e., no irrigation and infrequent mowing), ET was not significantly different for five perennial grasses. All grasses used more water than the bare-ground treatment. Soil water uptake was greatest in the upper soil layer (O to 25 cm) and decreased with depth. Few differences in water uptake were noted among grasses within each soil layer.