Emergence, Early Growth, and Community Dynamics of Key Rice Weeds in California in Variable Irrigation Systems: Impacts on System Productivity and Yield Losses.

摘要

Rice is one of our most important crops, providing 20% of global annual consumed calories. It is the staple grain for about half of the world's population. Grown in a variety of agroecosystems, ranging from lowland irrigated to upland rainfed, it is second only to wheat in hectares planted worldwide. Irrespective of location or farming system, weeds are the greatest biological constraint to yield. Farmer inputs of both labor and capital towards weed management are high, and are expected to increase as herbicide resistance continues to spread worldwide.;In California, rice is grown on approximately 200,000 hectares in the state's central valley. The majority of the acreage has been grown under continuously flooded irrigation since the start of commercial rice culture in the state in 1912, and as a result, weeds have become highly adapted to this anaerobic system. Herbicide resistance was first discovered in 1993 in two weed species with resistance to acetolactate synthase (ALS) inhibiting herbicides. Since then, another nine species have evolved resistance to an additional four modes of action. Coupled with the spread of herbicide resistance is increasing pressure to reduce water usage, both due to a statewide drought, as well as from the public, which perceives rice to be a heavy water user.;The overall objective of the first study was to understand the impacts of alternative irrigation systems on weed community dynamics. Three irrigation systems were evaluated: 1) a conventional water-seeded control (WS-Control) with a permanent flood of 10-15 cm from planting until one month prior to harvest; 2) WS-alternate wet and dry (WS-AWD) which was flooded from planting until canopy closure after which flood water was allowed to subside and the field re-flooded when the soil volumetric water content (VWC) reached 35%; and 3) drill-seeded AWD (DS-AWD) where rice was drill-seeded, then flush irrigated to establish the crop after which it was flooded until canopy closure and then underwent AWD similar to WS-AWD. The results indicate that there are significant changes in weed community composition both between irrigation systems as well as within irrigation systems over time. The most significant impacts are the increase in smallflower umbrella sedge (Cyperus difformis L.) populations under the WS-AWD system when compared to the continuously flooded system, as well as the large increase in the watergrass species ( Echinochloa (L.) Beauv. spp.) under the DS-AWD system. Since both smallflower umbrella sedge and watergrass are known to have herbicide resistant populations in California, this may restrict the adoption of the proposed alternative irrigation systems by farmers.;The second study evaluates the predictive capability of two yield loss models in naturally occurring mixed-weed species infestations in rice. The models were calibrated at the California Rice Experiment Station in 2013-2014, and then validated over two years at four sites. A relative cover assessment at rice canopy closure was used as the independent variable, and the weed species groups evaluated were the grasses, sedges, broadleaves, and all weeds combined. At the calibration site, grass cover alone was the best predictor in 2013, and grasses and sedge cover combined was the best predictor in 2014. Results of the model validation indicate that the relative cover of grasses and sedges combined was the best predictor of rice yield losses in five out of seven site-years.;The third study modifies a previously-developed population-based threshold model (PBTM) for multiple-herbicide resistant and susceptible late watergrass (Echinochloa phyllopogon (Stapff). Koss) to better predict emergence and early growth to the 1-leaf stage in rice under both continuously flooded and intermittent irrigation conditions. The model was modified using data on emergence under four burial depths (0.5 cm, 2 cm, 4 cm, and 6 cm) and three irrigation systems: 1) Daily flush (saturated soil); 2) Continuously flooded (anaerobic conditions); and 3) Intermittent flush (water-stressed conditions). Emergence of the resistant population was greater under the water-stressed conditions and emergence of both populations was negatively correlated with seed burial depth. Model fit of the field validation was best under the continuously flooded system, but the observed emergence in the field was biphasic, indicative of two separate emergence events. The model from this study only predicted the first emergence flush.