Weed management remains a high priority for organic farmers, whose fieldsgenerally have higher weed density and species diversity than those of theirconventional counterparts. We explored whether variability in farmerknowledge and perceptions of weeds and weed management practices werepredictive of variability in on-farm weed seedbanks on 23 organic farms innorthern New England. We interviewed farmers and transcribed and codedinterviews to quantify their emphasis on concepts regarding knowledge ofecological weed management, the perceived risks and benefits of weeds, andthe perceived risks and benefits of weed management practices. Tocharacterize on-farm weed seedbanks, we collected soil samples from fivefields at each farm (115 fields total) and measured germinable weed seeddensity. Mean weed seed density per farm ranged from 2,775 seeds m−2 to 24,678 seeds m−2 to a soil depth of 10 cm.Farmers most often reported hairy galinsoga and crabgrass species (Digitaria spp.) as their most problematic weeds. Theproportion of the sum of these two most problematic weeds in each farm'sseedbank ranged from 1 to 73% of total weed seed density. Farmer knowledgeand perceptions were predictive of total seed density, species richness, andproportion of hairy galinsoga and crabgrass species. Low seed densities wereassociated with farmers who most often discussed risks of weeds, benefits ofcritical weed-free management practices, and learning from their ownexperience. These farmers also exhibited greater knowledge of managing theweed seedbank and greater understanding of the importance of a long-termstrategy. Targeted education focusing on this set of knowledge and beliefscould potentially lead to improved application and success of ecologicalweed management in the future, thus decreasing labor costs and timenecessary for farmers to manage weeds.

The use of eucalyptus leaves for weed control in maize-based croppingsystems is proposed. Aqueous extracts of eucalyptus are known to exertphytotoxicity on many weeds and crops, but there is also experimentalevidence of the relative tolerance of maize. Based on in vitro dose-responsebioassays of leaf aqueous extracts, we conducted greenhouse pot experimentstesting incorporated eucalyptus leaves as green manure. The phytotoxiceffects were tested on the germination, establishment and growth of maizeand some representative accompanying weeds, in comparison to the PREherbicide metolachlor. Eucalyptus fresh leaves incorporated into the soil asgreen manure at 1 and 2% w/w reduced the emergence of the dicot weed speciesredroot pigweed and black nightshade. After one month of incorporation, bothdoses reduced aerial biomass >94% two monocot weed species (barnyardgrass+ large crabgrass) with respect to the eucalyptus-free pots, and around 80%for the small seeded dicots. Although the aerial biomass of maize wasreduced by 33%, the final relative yield of maize biomass with respect tothe untreated control increased by 37%. On the assessment of the temporalphytotoxic effects, the reduction of aerial biomass in maize could beovercome by adopting a relay-planting of maize after 12 to 15 days fromeucalyptus incorporation. Our results constitute evidence that theincorporation of E. globulus residues to soil could be afeasible practice to reduce the reliance on synthetic herbicides inmaize-based cropping systems.

Predicting weed emergence dynamics can help farmers to plan more effectiveweed control. The hydrothermal time concept has been used to model emergenceas a function of temperature and water potential. Application of thisconcept is possible if the specific biological thresholds are known. Thisarticle provides a data set of base temperature and water potential of eightmaize weeds (velvetleaf, redroot pigweed, common lambsquarters, largecrabgrass, barnyardgrass, yellow foxtail, green foxtail, and johnsongrass).For five of these species, two ecotypes from two extreme regions of thepredominant maize-growing area in Italy (Veneto and Tuscany), were collectedand compared to check possible differences that may arise from using thesame thresholds for different populations. Seedling emergence of velvetleafand johnsongrass were modeled using three different approaches: (1) thermaltime calculated assuming 5 C as base temperature for both species; (2)thermal time using the specific estimated base temperatures; and (3)hydrothermal time using the specific, estimated base temperatures and waterpotentials. All the species had base temperatures greater than 10 C, withthe exception of velvetleaf (3.9 to 4.4 C) and common lambsquarters (2.0 to2.6 C). All species showed a calculated base-water potential equal or up to−1.00 MPa. The thresholds of the two ecotypes were similar for all thestudied species, with the exception of redroot pigweed, for which the Venetoecotype showed a water potential lower than −0.41 MPa, whereas it was −0.62MPa for the Tuscany ecotype. Similar thresholds have been found to be usefulin hydrothermal time models covering two climatic regions where maize isgrown in Italy. Furthermore, a comparison between the use of specific,estimated, and common thresholds for modeling weed emergence showed that,for a better determination of weed control timing, it is often necessary toestimate the specific thresholds.