In the absence of tillage, seeds infiltrate into an agricultural soil via cracks, the activities of soil fauna, and frost action. This infiltration is slow. For example, Moss (1985b) found 92% of Alopecurus myosuroides seeds in the top 2.5 cm of soil 10 months after sowing. After 34 months, 78% were still in the top 2.5 cm. Other studies show similar results (Weaver & Cavers 1979; Van Esso, Ghersa & Soriano, 1986).
Tillage implements move seeds vertically to different extents. Most studies of seed movement have observed the distribution of seeds or beads that were sown on the soil surface and then tilled in with one or more tillage operations (Pawlowski & Malicki, 1968; Moss, 1988; Staricka et al., 1990; Yenish et al., 1996). Others have compared the distribution of natural seed populations before and after tillage (Russel & Mehta, 1938; Roberts, 1963; Wicks & Somerhalder, 1971; Yenish, Doll & Buhler, 1992). Hulburt & Menzel's (1953)
measurements of how surface-applied 32P was mixed by disking and rototil-ling may indicate the way surface seeds are moved by these operations.
Typically, a single moldboard plowing of surface seeds results in a skewed, bell-shaped distribution of seed density with depth (Figure 4.3). The direction and degree of skewing has varied among studies (Mohler, 1993), perhaps due to variation in depth of operation and degree of soil turning by the plows. A variety of factors affect the degree of inversion. These include design of the moldboard and coulters, speed of operation, amount of roots and crop residue present, and soil structure, texture, and moisture.
In contrast with moldboard plowing, a single operation with other implements produces a monotonic decline in seed density (Figure 4.3). A second operation (usually the following season) makes the distribution more uniform with all implements (Figure 4.3), but seeds still tend to be concentrated near the surface with tine and disk type implements.
Using colored beads placed at different depths, Cousens & Moss (1990) and Grundy, Mead & Burston (1999) developed data on movement to and from multiple layers of the soil and then used these data to model seed distributions after multiple tillage events. The distributions predicted by Cousens & Moss (1990) for a single cohort of seeds moved repeatedly by a single type of implement were similar to seed distributions after multiple tillage events observed in empirical studies (Figure 4.3) (Rottele & Koch, 1981; Van Esso, Ghersa & Soriano, 1986). Cousens & Moss (1990) also considered the situation where weeds shed seeds onto the surface each year. The stable distributions predicted under the assumptions of the model after many cycles were similar to those for a single cohort after one tillage operation (Figure 4.4). Published seed distributions after many years of moldboard plowing with seed input (Roberts & Stokes, 1965; Fay & Olson, 1978) are more uniform than predicted by Cousens & Moss (1990). In real agricultural fields, the distribution of seeds is rarely stable because years of high seed input usually occur sporadically among years of successful control and little seed input. Following one or more high-input years, the distribution would look something like those in Figure 4.3 for one tillage operation. After one or more years of good weed control, the distribution would become more uniform.
Was this article helpful?