Spatial distribution of enhanced atrazine degradation across northeastern Colorado cropping systems.

Reports of enhanced atrazine degradation and reduced residual weed control have increased in recent years, sparking interest in identifying factors contributing to enhanced atrazine degradation. The objectives of this study were to (i) assess the spatial distribution of enhanced atrazine degradation in 45 commercial farm fields in northeastern Colorado (Kit Carson, Larimer, Logan, Morgan, Phillips, and Yuma counties) where selected cultural management practices and soil bio-chemo-physical properties were quantified; (ii) utilize Classification and Regression Tree (CART) Analysis to identify cultural management practices and (or) soil bio-chemophysical attributes that are associated with enhanced atrazine degradation; and (iii) translate our CART Analysis into a model that predicts relative atrazine degradation rate (rapid, moderate, or slow) as a function of known management practices and (or) soil properties. Enhanced atrazine degradation was widespread within a 300-km radius across northeastern Colorado, with approximately 44% of the fields demonstrating rapid atrazine degradation activity (laboratory-based dissipation time halflife [DT50] < 3 d). The most rapid degradation rates occurred in fields that received the most frequent atrazine applications. Classification and Regression Tree Analysis resulted in a prediction model that correctly classified soils with rapid atrazine DT50 80% of the time and soils with slow degradation (DT50 > 8 d) 62.5% of the time. Significant factors were recent atrazine use history, soil pH, and organic matter content. The presence/absence of atzC polymerase chain reaction (PCR) product was not a significant predictor variable for atrazine DT50. In conclusion, enhanced atrazine degradation is widespread in northeastern Colorado. If producers know their atrazine use history, soil pH, and OM content, they should be able to identify fields exhibiting enhanced atrazine degradation using our CART Model.