Sustainable invasive weed management must address treatment effects on desired vegetation. Our objective was to determine the influence of clopyralid plus 2,4-D, glyphosate, and fosamine, at various application rates and timing, on the density and biomass of Russian knapweed and desired plant groups growing in association with this invasive weed. In a randomized complete block design with four replications, three herbicides by three herbicide rates by three herbicide application timings and a nontreated control were factorially applied to two sites located along the Missouri River riparian corridor in Montana. Clopyralid plus 2,4-D, glyphosate, and fosamine were applied during the spring rosette stage of Russian knapweed (June), the bud to bloom stage of Russian knapweed (July), or the flowering stage of Russian knapweed (August). Herbicide rates were considered low, medium, and high based on label rates of clopyralid plus 2,4-D, glyphosate, or fosamine. Density and biomass of all species were sampled 3 yr after treatment. Russian knapweed biomass decreased from 125 to about 25 g/m2 using clopyralid plus 2,4-D, irrespective of rate or timing of application. Russian knapweed density was reduced by about half by this mixture of herbicides. Nonnative grass density and biomass were maintained, whereas native grasses increased using clopyralid plus 2,4-D at medium or high rates. Neither glyphosate nor fosamine provided substantial Russian knapweed control or increases in grasses. Too few forbs were present to analyze their response to the treatments. We believe that herbicides must be combined with revegetation in areas lacking a diverse mixture of desired species capable of capturing resources made available by controlling Russian knapweed.

In the mid-Atlantic region, there is increasing interest in the use of intercropping strategies to establish cover crops in corn cropping systems. However, intercropping may be limited by potential injury to cover crops from residual herbicide programs. Field experiments were conducted from 2013 to 2015 at Pennsylvania, Maryland, and New York locations (n=8) to evaluate the effect of common residual corn herbicides on interseeded red clover and annual ryegrass. Cover crop establishment and response to herbicide treatments varied across sites and years. S-metolachlor, pyroxasulfone, pendimethalin, and dimethenamid-P reduced annual ryegrass biomass relative to the nontreated check, whereas annual ryegrass biomass in acetochlor treatments was no different compared with the nontreated check. The rank order of observed annual ryegrass biomass reduction among chloroacetamide herbicides was S-metolachlor>pyroxasulfone>dimethenamid-P>acetochlor. Annual ryegrass biomass was not reduced by any of the broadleaf control herbicides. Mesotrione reduced red clover biomass 80% compared to the nontreated check. No differences in red clover biomass were observed between saflufenacil, rimsulfuron and atrazine treatments compared to the nontreated check. Red clover was not reduced by any of the grass control herbicides. This research suggests that annual ryegrass and red clover can be successfully interseeded in silt loam soils of Pennsylvania following use of several shorter-lived residual corn herbicides, but further research is needed in areas with soil types other than silt loam or outside of the mid-Atlantic cropping region.

Sericea lespedeza is an invasive weed in the tallgrass prairies of the Southern Great Plains. Field experiments were initiated in 1995 at three locations in central Oklahoma to evaluate control of sericea lespedeza for several years after treatment with herbicide. Herbicide treatments included triclopyr at 560 and 840 g ae/ha, fluroxypyr at 210 and 560 g ae/ha, and metsulfuron at 13 and 21 g ai/ha applied at simple-stem (SS), branched-stem (BS), and flowering (FL) growth stages of sericea lespedeza. At all three locations, applications of triclopyr and fluroxypyr at the BS growth stage resulted in less than 4% of the pretreatment sericea lespedeza stem density remaining in the first growing season after treatment (GSAT). Metsulfuron applied at the FL growth stage resulted in 0 to 9% of the pretreatment stem density remaining in the first GSAT. Regardless of rate, triclopyr and fluroxypyr applied at the BS growth stage provided the most consistent long-term control of sericea lespedeza. Percentage of pretreatment stem density remaining with these treatments was 0 to 20% at two locations in the third GSAT, and 4 to 15% at one location in the fifth GSAT. Aboveground biomass yields of desirable grasses (bermudagrass, indiangrass, and little bluestem) at two locations were greater than that of the untreated check in the second and third GSAT in all herbicide treatments. But biomass yield of bermudagrass did not increase when sericea lespedeza was controlled at the location with eroded soil conditions.

Field studies were conducted in 2007 and 2008 at seven sites in Ohio, Indiana, and Illinois to determine the effect of PRE herbicide and POST application timing on weed control and yield of glyphosate-resistant corn. Levels of PRE herbicide included none; low—atrazine; medium—atrazine and metolachlor; and high—atrazine, mesotrione, and metolachlor. Glyphosate was applied POST when corn was 30 cm tall, or 1 or 2 wk later. Common lambsquarters, giant foxtail, and giant ragweed infested at least six of the seven sites, and other weed species occurred at two to three sites. Control of weeds at the time of POST application ranged from 48 to 91%, 58 to 99%, and 87 to 100% for the low, medium, and high levels of PRE herbicide, respectively, averaged over POST application timing. Control of giant foxtail and redroot pigweed decreased by about 20% between the second and third POST timing, averaged over PRE herbicide, but control of other weeds was similar among timings. Late-season control of common ragweed, velvetleaf, common lambsquarters, and Pennsylvania smartweed exceeded 90%, regardless of PRE herbicide or POST timing. Control of redroot pigweed, ivyleaf morningglory, and giant ragweed was as low as 74, 67, and 83%, respectively, but the high level of PRE herbicide resulted in 90 to 97% control of these weeds. An interaction between PRE herbicide and POST timing for late-season control of giant foxtail, tall waterhemp, and yellow nutsedge reflected the more effective control among POST timings from the higher levels of PRE herbicide. The overall trend in this study was for more effective weed control in PRE/POST herbicide programs with more comprehensive PRE herbicides that have substantial activity on both grass and broadleaf weeds. Highest yield occurred where the PRE treatment consisted of a two- or three-way combination of herbicides applied at 50% of the recommended rate or higher. Yield was reduced at all POST timings with atrazine alone or in the absence of PRE herbicide.

Exotic species invasions in natural areas are one of the most significant threats to biological diversity globally. Pest plants pose a significant problem because they often go undetected until widespread ecological damage has already occurred. Effective control is both uncertain and expensive. However, not all introduced species become invasive, leading to the hope that we can develop risk assessment criteria for new plant introductions. Two recently proposed assessment programs are reviewed, one based on North American woody plants and the other based on Australian pest species, and the challenges in their application are discussed. Among the significant issues are spatial and temporal variation in plant performance that affect the documentation of invasive behavior and the tendency for horticulturists to value traits that produce invasive behavior (rapid growth, early and consistent flowering, lack of pests and diseases, and vegetative persistence). Two policy alternatives are suggested for botanical gardens as examples of models for plant introduction policies that could be adapted to other institutions: the Conservation Aware Garden and the Strict Conservation Garden. The former is based on risk assessment, whereas the latter prohibits movement of species across barriers to their dispersal. Information needs, the importance of international communication, and adaptive management are discussed as elements of a program to reduce the spread of pest invaders.

Information on the effects of multiple weed management tactics in corn is needed to develop integrated weed management systems. The effectiveness and compatibility of an in-row cultivator as compared to a standard interrow cultivator used with reduced rates of a soil-applied herbicide, rotary hoeing, and/or a bioeconomic model for POST herbicide selection was examined. Weed control with a single rotary hoeing at corn emergence controlled annual weeds similarly to two rotary hoeings. One-third recommended use rate of alachlor controlled weeds similarly to a two-thirds rate. Reduced rates of alachlor controlled more weeds than rotary hoeing over 2 yr. The in-row cultivator required early-season weed control (rotary hoeing or reduced alachlor rate) for optimum efficacy. The in-row cultivator provided better weed control than the standard cultivator while the cost of operating the two cultivators was similar. Thus, the in-row cultivator was more efficient than the standard cultivator. Furthermore, less intensive early-season weed control was required with the in-row cultivator for maximum weed control as compared to the standard cultivator. Rotary hoeing plus the in-row cultivator provided similar weed control to other weed management tactics that required both soil-applied and POST herbicides. Gross margin was influenced more by corn yield than cost of weed management tactics.

In this research, the deep-learning optimizers Adagrad, AdaDelta, Adaptive Moment Estimation (Adam), and Stochastic Gradient Descent (SGD) were applied to the deep convolutional neural networks AlexNet, GoogLeNet, VGGNet, and ResNet that were trained to recognize weeds among alfalfa using photographic images taken at 200×200, 400×400, 600×600, and 800×800 pixels. An increase in the image sizes reduced the classification accuracy of all neural networks. The neural networks that were trained with images of 200×200 pixels resulted in better classification accuracy than the other image sizes investigated here. The optimizers AlexNet and GoogLeNet trained with AdaDelta and SGD outperformed the Adagrad and Adam optimizers; VGGNet trained with AdaDelta outperformed Adagrad, Adam, and SGD; and ResNet trained with AdaDelta and Adagrad outperformed the Adam and SGD optimizers. When the neural networks were trained with the best-performing input image size (200×200 pixels) and the best-performing deep learning optimizer, VGGNet was the most effective neural network, with high precision and recall values (≥0.99) when validation and testing datasets were used. Alternatively, ResNet was the least effective neural network in its ability to classify images containing weeds. However, there was no difference among the different neural networks in their ability to differentiate between broadleaf and grass weeds. The neural networks discussed herein may be used for scouting weed infestations in alfalfa and further integrated into the machine vision subsystem of smart sprayers for site-specific weed control.

Field studies were conducted at Yoakum and Stephenville, TX; Jay, FL; and Midville and Plains, GA, to determine the persistence of imazapic applied to peanuts at 0, 70, 140, and 210 g ai/ha. The following year, cotton, sorghum, and corn were planted in the treated plots in Texas, cotton was planted in Florida, and corn and cotton were planted in Georgia and evaluated for carryover injury. Data collected to determine injury included plant heights and weights. In 1999 in Texas and in Florida and Georgia, there was no significant carryover injury to rotational crops from any of the imazapic rates. Data on cotton and sorghum plant height from Texas in 2000 showed height reductions for the 210-g/ha rate on cotton and the 140- and 210-g/ha rates on sorghum. These data showed no significant carryover effects to rotational crops from the 70-g/ha rate of imazapic applied to peanuts the previous year.

Studies were conducted at four sites during a 2-yr period in Oklahoma, Texas, and Arkansas to determine effectiveness and safety of halosulfuron in honeydew crops. Halosulfuron applied postemergence at 26.3 to 78.8 g ai/ha controlled yellow nutsedge 85 to 97%, golden crownbeard 100%, and tumble pigweed 83 to 95%. Control of yellow nutsedge continued to increase for 3 to 6 wk after treatment. Golden crownbeard and tumble pigweed efficacy increased to its highest levels after 4 and 3 wk, respectively. Reduced crop growth and yellowing of foliage did not exceed 13%. No differences were recorded for yield, earliness, or percentage of marketable fruit.

A field experiment was conducted over 2 yr to study efficacy of bromoxynil mixtures with pyrithiobac or MSMA applied postemergence (POST) with and without fluometuron or fluometuron plus pendimethalin preemergence (PRE) for control of broadleaf weeds in ultra narrow row bromoxynil-resistant cotton in the Mississippi Delta. Bromoxynil applied POST (single or sequential) provided variable control of common purslane (<9%), sicklepod (<35%), Palmer amaranth (<46%), prickly sida (>75%), hyssop spurge (>79%), hemp sesbania (>96%), and pitted morningglory (100%) at 4 wk after early POST (WAT). Broadleaf weed control increased when PRE herbicides were followed by bromoxynil or bromoxynil plus pyrithiobac or MSMA POST. Weed control generally decreased at harvest compared to 4 WAT, and the decrease was greater in bromoxynil POST-only programs compared to bromoxynil POST following PRE programs. Seed cotton yield with bromoxynil POST-only programs was lower (400 to 2,810 kg/ha) compared to bromoxynil POST programs following PRE herbicides (2,150 to 3,720 kg/ha). Early-season weed interference and variable control of weeds in bromoxynil POST-only programs resulted in greater cotton stand reduction and lower open bolls per plant compared to bromoxynil POST programs following PRE herbicides.

Flower production and pollen dispersal patterns of diclofop-methyl-susceptible and -resistant Italian ryegrass were examined to determine whether evolution of herbicide resistance could be controlled by cross-fertilization between biotypes. The susceptible and resistant ryegrass biotypes differed in timing and abundance of both ovule production and pollen release such that pollen from the susceptible plants had a much greater chance of fertilizing the resistant plant population than vice versa. Susceptible ryegrass, growing with or without wheat competition, produced more than 60% of its seeds before any pollen from resistant plants was released. In contrast, throughout the course of resistant plant seed generation, pollen from susceptible plants composed at least 30% of the total pollen load. These phenological differences, in conjunction with herbicide treatment results for third-generation ryegrass seedlings, suggest that gene flow can be used to reduce development of diclofop-methyl resistance in ryegrass populations within wheat cropping systems. In a wheat field infested with resistant ryegrass, cessation of herbicide application and sowing of a susceptible ryegrass biotype can be expected to reduce the evolution of diclofop-methyl resistance by at least 6% per year.

Field experiments were conducted to evaluate control of 90- to 100-cm-tall ragweed parthenium in a noncropped situation in Haryana State, India, during 2000 and 2001. Atrazine, 2,4-D ethyl ester, atrazine plus 2,4-D, metribuzin, metsulfuron, chlorimuron, glufosinate with and without surfactant, glyphosate with and without surfactant, and glyphosate formulations MON 8793 and 8794 were sprayed on ragweed parthenium. Also, the effect of water quality was studied with flat-fan and flood-fan nozzles using glyphosate and its formulation MON 8793 against ragweed parthenium and associated weeds. Glyphosate MON 8793 and 8794 at 3.6 kg ae/ha provided excellent control of ragweed parthenium followed by glyphosate at 2.7 or 5.4 kg/ha, with no recovery until 18 wk after treatment (WAT). Addition of 0.1% v/v surfactant (MON 0818) to glyphosate at 2.7 kg/ha provided similar control to that of glyphosate alone at 5.4 kg/ha. Other herbicides failed to provide satisfactory control of ragweed parthenium. In the water quality study, glyphosate at 2.7 and 5.4 kg/ha and glyphosate MON 8793 at 2.7 and 3.6 kg/ha provided similar control of ragweed parthenium at 18 WAT. Glyphosate was antagonized less by tap water (0.45 mM Ca) than canal (0.7 mM Ca) and hand-pump water (1 mM Ca). Neither glyphosate nor glyphosate MON 8793 provided good control of purple nutsedge, velvetleaf, garden spurge, threelobe false mallow, jimsonweed, giant milkweed, Indian jujube, or tropical spiderwort, but crowfootgrass, green foxtail, sprawling signalgrass, and spiny amaranth were controlled. Glyphosate at 5.4 kg/ha and glyphosate MON 8793 at 3.6 kg/ha provided more than 80% control of bermudagrass at 8 WAT, which was significantly better than the 2.7 kg/ha rate. Flat-fan nozzles provided better efficacy of applied herbicides than flood-fan nozzles at 4 WAT on ragweed parthenium.

Field experiments were conducted to determine effects of adding 2.8 kg/ha (NH4)2SO4 and/or substituting BCH 81508 S for crop oil with sethoxydim applied to large crabgrass, corn, johnsongrass, and common bermudagrass. Adding (NH4)2SO4 did not affect control of johnsongrass or common bermudagrass but did increase control of large crabgrass and corn by 6 and 15%, respectively. Cora control was enhanced more by (NH4)2SO4 at lower sethoxydim rates. Substituting BCH 81508 S for crop oil increased seedling johnsongrass, large crabgrass, common bermudagrass, rhizome johnsongrass, and corn control by 7, 9, 9, 10, and 15%, respectively. A sethoxydim rate by spray adjuvant interaction was observed with seedling johnsongrass, common bermudagrass, and corn where response to BCH 81508 S decreased as sethoxydim rates increased. The effects of substituting BCH 81508 S for crop oil and adding (NH4)2SO4 were independent.

We conducted a field experiment in 2007 and 2008 and repeated it in a separate field in 2008 and 2009 to test the effectiveness of two herbicides applied at two different times on weed control during switchgrass establishment. At 4 and 6 wk after switchgrass planting (WAP), sulfosulfuron was applied at 0.04 and 0.07 kg ai ha−1 with nonionic surfactant and quinclorac was applied at 0.28, 0.42, and 0.56 kg ai ha−1 with methylated seed oil. Herbicides applied at 4 WAP tended to be more effective than at 6 WAP. Sulfosulfuron provided greater control of smooth pigweed; however, quinclorac provided greater control of yellow foxtail, a grass weed that has traditionally been difficult to control with herbicides in switchgrass. Average yellow foxtail control was 73, 87, and 93% for quinclorac at 0.28, 0.42, and 0.56 kg ha−1, respectively, compared to 62 and 60% for sulfosulfuron at 0.04 and 0.07 kg ha−1, respectively. Switchgrass injury (chlorosis and height reduction relative to the untreated control) was observed, but most symptoms were not detectable by 8 wk after treatment (WAT) for most treatments. Plots that received quinclorac at 0.56 kg ha−1 at 6 WAP tended to have relatively low weed biomass and high total aboveground yield in the establishment year and relatively high total aboveground yield in the year after establishment.

Biologically active natural products of microbial origin are attractive candidates for possible use in agriculture. They may be obtained by fermentation, used in their natural state, or subjected to synthetic modification for specific uses. These natural products are characterized by high specific activity and high selectivity, and they are biodegradable. The structures are extremely diverse and represent many classes of compounds ranging from very complex to simple. Cyclocarbamide A and B, from Streptoverticillium sp., have marked preemergence herbicidal activity. Nigerazine A and B, from Aspergillus niger van Tieghem, also inhibit root growth in certain plants. Citreoviridin, from Penicillium charlesii Smith, preferentially controls the growth of monocotyledonous plants, as does a synthetic derivative of cladosporin, from Aspergillus repens DeBary, which bleaches chloroplasts. The 12-membered fungal macrolides (macrocyclic lactones) also inhibit root growth in many test plants and offer templates for further synthetic work. Herbicidins, from Streptomyces saganonensis, are particularly effective against barnyardgrass, goosegrass, tufted mannagrass, and green panicum. Alternaria eichorniae Nag Raj et Ponnappa produces a toxin that is active against waterhyacinth and represents one of the more exotic structures. The macrocyclic trichothecenes are a significant class of natural products that tend to concentrate against a gradient in seeds of certain plants, which resist these microbially derived metabolites thereby producing seed with “built-in” natural herbicides.

Few herbicides are available for weed control in vegetable production systems using low-density polyethylene (LDPE) plastic mulch. With the elimination of methyl bromide for pest management and subsequent use of various alternative fumigants, the need for herbicides in vegetable production systems has increased. An experiment was conducted to evaluate tolerance of transplant summer squash and tomato to carfentrazone, flumioxazin, glyphosate, halosulfuron, or paraquat applied to the mulch prior to transplanting. After applying herbicides overtop of the mulch but prior to vegetable transplant, the mulch was either irrigated with 1.0 cm of water or not irrigated. Carfentrazone did not affect either crop regardless of irrigation. Irrigation readily removed glyphosate and paraquat from the mulch, as there was no adverse crop injury in these treatments. In the absence of irrigation, glyphosate and paraquat reduced squash diameter and tomato heights 18 to 34% at 3 wk after transplanting (WAT). Squash and tomato fruit numbers and fruit biomass (yield) were reduced 17 to 37%, and 25 to 33%, respectively. Halosulfuron reduced squash diameter and yield 71 to 74% and tomato heights and yields 16 to 37% when mulch was not irrigated prior to transplanting. After irrigating, halosulfuron had no affect on tomato, but reduced squash growth and yield 40 to 44%. Flumioxazin killed both crops when the mulch was not irrigated; and reduced squash yield 56% when irrigated. With irrigation, flumioxazin did not impact tomato fruit number, but did reduce tomato weight by 25%. These studies demonstrate the safety of carfentrazone, applied on mulch prior to transplanting either squash or tomato, regardless of irrigation, and also demonstrate the safety of glyphosate and paraquat if irrigated prior to transplanting. Conversely, flumioxazin should not be applied over mulch before transplanting either crop, regardless of irrigation. Halosulfuron application over mulch should be avoided before transplanting squash, regardless of irrigation, but can be applied prior to transplanting tomato if irrigated.

Field studies were conducted in 2001 and 2002 to determine the effect of tillage on the emergence of common waterhemp from the soil seedbank. Emergence of common waterhemp was three times greater in no-till than chisel-till cultivation. Tillage did not affect the initial time of emergence; however, the time to 50% emergence was longer in no-till than chisel till. Duration of emergence did not differ among tillage systems. Common waterhemp seed was concentrated near the soil surface in no-till plots, whereas seed in the chisel-till plots were primarily found between 9 and 15 cm. The delayed and increased emergence in no-till cultivation may contribute to the greater problems in managing common waterhemp in no-till plots compared with plots where tillage is used.

Field research was conducted for 3 yr to evaluate crop response and weed control under conventional and reduced tillage in drill- and water-seeded imidazolinone-tolerant (IT) rice culture. Imazethapyr was applied at 70 g ai/ha PRE followed by (fb) imazethapyr at 70 g/ha applied POST to three- to four-leaf rice or at 105 g/ha PRE fb 70 g/ha POST. In both conventional and reduced tillage systems, imazethapyr applied PRE fb POST at 70 g ai/ha controlled red rice, barnyardgrass, Amazon sprangletop, and rice flatsedge 87 to 99% 35 d after POST treatment (DAT). At 35 DAT, Indian jointvetch control with sequential applications of imazethapyr was as high as 70% in water-seeded rice but no more than 54% in drill-seeded rice. Tillage, seeding method, and imazethapyr rate had no effect on days to 50% heading, seeds per panicle, seed weight per panicle, or percentage of seed harvest. However, a reduction of 27% in days to 50% heading, 80% in seeds per panicle, 84% in seed weight per panicle, and 100% in percentage seed harvest index occurred when imazethapyr was not applied because of weed interference. Culm number was reduced 28%, and culm weight 32% under reduced tillage compared with conventional tillage. With sequential applications of imazethapyr at 70 g/ha, rice yield was 63% greater when rice was water-seeded compared with drill-seeded. No differences in tillage systems for weed control, days to 50% heading, seed number, seed weight per panicle, percent seed, panicle height, lodging, or yield were observed. Results of these experiments demonstrate imazethapyr will effectively control weeds in both water- and drill-seeded rice and that reduced tillage can be used without negatively affecting rice production.

Broadleaf weeds and nutsedge are persistent problems with limited management options for strawberry growers in Florida. Experiments were conducted in 2012 to 2013 (year 1) and 2013 to 2014 (year 2) at the Gulf Coast Research and Education Center in Wimauma, FL, to evaluate tolerance of two strawberry cultivars (‘Strawberry Festival' and ‘WinterStar'™) to PRE herbicides applied on the bed top or injected through the drip tape. The bed-top herbicides were applied to formed beds following fumigation but prior to laying the plastic mulch. Herbicides evaluated included two rates of EPTC, fomesafen, halosulfuron, and S-metolachlor as well as tank mixes of EPTC + S-metolachlor, fomesafen + S-metolachlor, and napropamide + oxyfluorfen. Drip-applied herbicides were applied through a single drip tape and strawberries were transplanted 1, 7, 15 and 30 d following application. Drip-applied herbicides included two rates of EPTC and fomesafen as well as two rates of a tank mix of napropamide + oxyfluorfen. None of herbicides applied on the bed top except S-metolachlor 214 g ai ha−1 consistently injured either strawberry cultivar or impacted yields. None of the drip-applied herbicides damaged the strawberry plants or reduced yields at all planting dates. In year 1, EPTC at 229 g ai ha−1 and EPTC at 229 g ha−1 + S-metolachlor at 107 g ha−1 resulted in a 70 to 86% reduction in purple nutsedge density compared to the nontreated control. No differences in nutsedge density among treatments were observed in year 2. The herbicides evaluated are safe for use on strawberry and would provide growers with alternative herbicide options, application timings, and application techniques. Further research is needed to evaluate efficacy.

Fruit trees in orchards of the mid-Atlantic region of the United States are often planted in vegetation-free rows alternating with grass alleys. Grass managed to suppress weeds but to compete minimally with fruit trees may be an alternative to herbicide and tillage. This research was conducted in the greenhouse and field to assess five different grasses that may suppress weeds without reducing yield of fruit trees. In the greenhouse with high seeding rates, red fescue competed more effectively than did chewings fescue, tall fescue, and perennial ryegrass with three weeds (damesrocket, cornflower, and chicory). However, with reduced seeding rates, similar to rates used in the field, grass competitiveness with weeds was similar between red fescue, tall fescue, and perennial ryegrass. Similar results were obtained during a 4-yr field experiment; roughstalk bluegrass competed least effectively with weeds but the other four grasses provided similar weed suppression—generally providing as much weed suppression as traditional herbicides. None of the candidate grasses significantly reduced yields of 10-yr-old apple and peach trees, although fruit size was affected by some grasses. The grass that was least suppressive of yield, roughstalk bluegrass, was the least effective in controlling weeds. Annual mowing in combination with four of the grasses tested is one option to manage the orchard floor with reduced herbicides, but fruit size may decrease.