The species composition and density of weed seed in the soil vary greatly and are closely linked to the cropping history of the land. Altering tillage practices changes weed seed depth in the soil, which plays a role in weed species shifts and affects efficacy of control practices. Crop rotation and weed control practices also affect the weed seedbank. Information on the influence of cropping practices on the weed seedbank should be a useful tool for integrated weed management. Decision aid models use information on the weed seedbank to estimate weed populations, crop yield loss, and recommend weed control tactics. Understanding the light requirements of weed seed may provide new approaches to weed management. Improving and applying our understanding of weed seedbank dynamics is essential to developing improved weed management systems. The principles of plant ecology must be integrated with the science of weed management to develop strategies that take advantage of basic plant responses in weed management systems for agronomic crops.

Emergence characteristics, including initial time of emergence, magnitude of emergence, and mean time of emergence, ofAmaranthus rudisSauer,Setaria faberiHerrm.,Eriochloa villosa(Thunb.) Kunth, andAbutilon theophrastiMedik. were studied in central Iowa during the first 3 yr following burial of seed collected and buried in the fall of 1994 or 1995. Although the initial emergence date varied among years, the emergence sequence among species was consistent.Eriochloa villosaand A.theophrastiwere the first species to emerge, with initial emergence occurring between April 28 and May 10. Initial emergence dates for these species were the same, except for 1995 whenA. theophrastiemerged 4 d prior toE. villosa. Amaranthus rudiswas the last species to emerge, with initial emergence ranging from 5 to 25 d afterA. theophrasti.First-year emergence in 1995 was 8% forA. theophrasti, 7% for A.rudis, 41% forE. villosa, and 33% forS. faberi, based on the number of buried seed. Proportional emergence during the first year following burial in 1996 was similar to 1995 forA. theophrastiandS. faberi, but greater emergence was observed in 1996 forE. villosaandA. rudis.During the 3 yr of both studies, cumulative emergence of the two grass species (43 to 71%) was higher than for the broadleaf species (13 to 35%). A high percentage of the total annual emergence ofE. villosaoccurred within the first 2 wk of initial emergence, whereas a high percentage ofA. rudisemergence occurred late in its emergence period. Emergence characteristics of the four species were consistent among years and seed sources.

First- and second-year seedbank emergence of 23 summer annual weed species common to U.S. corn production systems was studied. Field experiments were conducted between 1996 and 1999 at the Iowa State University Johnson Farm in Story County, Iowa. In the fall of 1996 and again in 1997, 1,000 seeds for most species were planted in plastic crates. Seedling emergence was counted weekly for a 2-yr period following seed burial (starting in early spring). Soil temperature at 2 cm depth was estimated using soil temperature and moisture model software (STM2). The Weibull function was fit to cumulative emergence (%) on cumulative thermal time (TT), hydrothermal time (HTT), and day of year (DOY). To identify optimum base temperature (Tbase) and base matric potential (ψbase) for calculating TT or HTT, Tbaseand ψbasevalues ranging from 2 to 17 C and −33 to −1,500 kPa, respectively, were evaluated for each species. The search for the optimal model for each species was based on the Akaike's Information Criterion (AIC), whereas an extra penalty cost was added to HTT models. In general, fewer seedlings emerged during the first year of the first experimental run (approximately 18% across all species) than during the second experimental run (approximately 30%). However, second-year seedbank emergence was similar for both experimental runs (approximately 6%). Environmental effects may be the cause of differences in total seedling emergence among years. Based on the AIC criterion, for 17 species, the best fit of the model occurred using Tbaseranging from 2 to 15 C with four species also responding to ψbase= −750 kPa. For six species, a simple model using DOY resulted in the best fit. Adding penalty costs to AIC calculation allowed us to compare TT and HTT when both models behaved similarly. Using a constant Tbase, species were plotted and classified as early-, middle-, and late-emerging species, resulting in a practical tool for forecasting time of emergence. The results of this research provide robust information on the prediction of the time of summer annual weed emergence, which can be used to schedule weed and crop management.

To improve understanding of relationships between crop diversity, weed management practices, and weed population dynamics, we used data from a field experiment and matrix models to examine how contrasting crop rotations affect velvetleaf. We compared a 2-yr rotation system (corn–soybean) managed with conventional rates of herbicides with a 4-yr rotation (corn–soybean–triticale + alfalfa–alfalfa) that received 82% less herbicide. In November 2002, a pulse of velvetleaf seeds (500 seeds m−2) was added to 7- by 7-m areas within replicate plots of each crop phase–rotation system combination. Velvetleaf seed, seedling, and reproductive adult population densities, seed production, and seed losses to predators were measured during the next year. Velvetleaf seed production was greater in the 4-yr rotation than in the 2-yr rotation (460 vs. 16 seeds m−2). Averaged over 12 sampling periods from late May to mid-November 2003, loss of velvetleaf seeds to predators also was greater in the 4-yr rotation than in the 2-yr rotation (32 vs. 17% per 2 d). Modeling analyses indicated that velvetleaf density in the 4-yr rotation should decline if cumulative losses of seeds produced in the soybean phase exceeded 40%. Achieving such a level of predation appears possible, given the observed rates of velvetleaf seed predation. In addition, no tillage occurs in the 4-yr rotation for 26 mo after soybean harvest, thus favoring seed exposure on the soil surface to predators. Models that included estimates of seed predation indicated that to prevent increases in velvetleaf density, weed control efficacy in soybean must be ≥ 93% in the 2-yr rotation, but could drop to 86% in the 4-yr rotation. These results support the hypothesis that diverse rotations that exploit multiple stress and mortality factors, including weed seed predation, can contribute to effective weed suppression with less reliance on herbicides.

Annual emergence and seed persistence of common waterhemp, velvetleaf, woolly cupgrass, and giant foxtail were characterized in central Iowa for 4 yr following burial of seeds collected and buried in autumn 1994. First-year emergence as a percentage of the original seed bank ranged from 5 to 40%, and the relative order was common waterhemp < velvetleaf < giant foxtail < woolly cupgrass. During the second and third years, there were no differences in percent emergence among species, with emergence percentages ranging from 1 to 9% of the original seed bank. During the fourth year, seedlings continued to emerge from only the velvetleaf and common waterhemp seed banks. A greater percentage of common waterhemp seed persisted each year and 12% of the original seed was recovered after 4 yr of burial. Five percent of the velvetleaf was recovered at the end of the fourth year. No woolly cupgrass and giant foxtail seed was recovered after the third and fourth years. The proportion of the seed that was accounted for from year to year through emergence and seed recovery varied by species and year. Total recovery of velvetleaf ranged from 61 to 87%, common waterhemp from 50 to 81%, woolly cupgrass from 29 to 79%, and giant foxtail from 23 to 79%. Based on the results of this research, velvetleaf and common waterhemp form more persistent seed banks than woolly cupgrass and giant foxtail. Therefore, woolly cupgrass and giant foxtail should be more amenable to management through seed bank depletion than velvetleaf and common waterhemp.