Path analysis, a statistical technique that differentiates between correlation and causation, was used to describe heavy metal adsorption by soil. A path analysis model that evaluates the effect of soil pH, CEC, organic carbon content (OC), and clay content on adsorption of Cd, Cu, Ni, Pb, and Zn by soils from two long‐term cropping experiments in Sutherland and Kanawha, IA, was investigated. In general, direct effects of soil properties on metal adsorption were pH > OC > CEC. Path analysis direct effects of pH on Cd (0.83), Cu (0.93), and Pb (0.96) by Galva (fine‐silty, mixed, mesic Typic Hapludoll) soils from the Sutherland experiment were highly significant (P < 0.01). Significant (P < 0.01) direct effects of pH on Cd (0.93), Cu (0.95), and Pb (0.85) adsorption by Webster (fine‐loamy, mixed, mesic Typic Haplaquoll) soils from the Kanawha experiment. Significant (P < 0.01) simple correlation coefficients were found between soil pH and Cd (0.64), Cu (0.89), and Pb (0.84) adsorption by Galva soils and between soil pH and Cd (0.82), Cu (0.74), and Pb (0.62) adsorption by Webster soils. Both path analysis and correlation analysis showed that soil pH was the predominant property affecting Cd, Pb, and Cu adsorption by soil. However, path analysis revealed three additional relationships not found by correlation analysis. They included direct effects of CEC on Ni adsorption (0.72*), OC on Pb adsorption (0.35*) by Galva soils, and OC on Cd adsorption (0.59*) by Webster soils. Partitioning of OC direct and indirect effects suggested that adsorption of Pb and Cd by soil organic matter through complexation reactions was more important than adsorption by organic matter CEC sites.

During 1985 to 1989, a series of field experiments were conducted at the Rice Research Station in Crowley, LA. Path analysis was employed to evaluate the competitive interaction between a weed (red rice) and cultivated rice (Mars). The path analysis quantified direct effects of red rice and Mars rice densities on the yield components (grain weight, percent filled florets, number of florets panicle−1, and panicles plant−1) of red rice and Mars rice. The model illustrated the direct and indirect effects of the yield components on fecundity and grain yield plant−1. The direct effects of Mars and red rice densities on panicles plant−1and florets panicle−1were always negative. In contrast, the effects of density on percent filled florets and grain weight varied from positive to negative and were relatively small, implying that they were determined primarily by density-independent factors. Path analysis indicated that the number of panicles plant−1and florets panicle−1were the most important yield components determining the responses of fecundity and grain yield to competition.

Four years of field experiments were used to assess the competitive ability of red rice, a weedy variety of rice, and ‘Mars’, a cultivar of rice. Red rice was the dominant competitor each of the 4 yr, and an average of one red rice plant reduced Mars grain yield per plant equal to the effect of four Mars plants. Intervarietal competition was more important than intravarietal competition for Mars, whereas the reverse was true for red rice. The product of the coefficients for intravarietal competition did not significantly exceed the product of the coefficients for intervarietal competition, indicating that the two varieties were competing for the same resources. Using the reciprocal yield model coefficients from 1989, grain yield losses of Mars, due to red rice densities of 4, 16, 25, and 300 plants m−2, were predicted to approximate 13, 37, 48, and 92%, respectively, at a fixed cultivar density of 100 plants m−2.

Herbicide contamination in surface runoff varies greatly depending on the herbicide management alternatives used. This study was conducted to determine if herbicide runoff is influenced by the timing of herbicide application with respect to maize (Zea mays L.) emergence. Preemergence (PRE) and postemergence (POST) applications of atrazine [6‐chloro‐N‐ethyl‐N′‐(1‐methylethyl)‐1,3,5‐triazine‐2,4‐diamine] were compared during 1988 and 1989 on a field with no prior history of atrazine use on a Barnes loam (fine‐loamy, mixed Udic Haploboroll) soil at the Swan Lake Research Farm near Morris, MN. A rainfall simulator applied precipitation events of 50 min with an intensity of 64 mm h−1 on Days 1,7, and 30 after PRE or POST applications of atrazine. Rainfall 1 d after application resulted in 32 g ha−1 of atrazine (less than 2% of the amount applied) in the runoff water and sediment for PRE treatments compared with 16 g ha−1 (less than 1% of the atrazine applied) for POST treatments. Similarly, 7 d after the herbicide applications, 20 g ha−1 was lost with PRE treatments and 1 g ha−1 for POST treatments. At 30 d, PRE treatments resulted in 8 g ha−1 being transported in the water and sediment compared with 1 g ha−1 for POST treatments. The rate of surface runoff, rather than the concentration of herbicide in the runoff, was the critical factor responsible for the reduction of the quantity of atrazine in the surface runoff with the POST treatments as compared with the PRE treatments.