Field experiments were conducted in Macon County, Georgia, during 2010 and 2011 to determine the impact of new herbicide-resistant cotton and respective herbicide systems on the control of glyphosate-resistant Palmer amaranth. Sequential POST applications of 2,4-D or glufosinate followed by diuron plus MSMA directed at layby (late POST-directed) controlled Palmer amaranth 62 to 79% and 46 to 49% at harvest when the initial application was made to 8- or 18–cm-tall Palmer amaranth, in separate trials, respectively. Mixtures of glufosinate plus 2,4-D applied sequentially followed by the layby controlled Palmer amaranth 95 to 97% regardless of Palmer amaranth height. Mixing glyphosate with 2,4-D improved control beyond that observed with 2,4-D alone, but control was still only 79 to 86% at harvest depending on 2,4-D rate. Sequential applications of glyphosate plus 2,4-D controlled Palmer amaranth 95 to 96% following the use of either pendimethalin or fomesafen. Seed cotton yield was at least 30% higher with 2,4-D plus glufosinate systems compared to systems with either herbicide alone. The addition of pendimethalin and/or fomesafen PRE did not improve Palmer amaranth control or yields when glufosinate plus 2,4-D were applied sequentially followed by the layby. The addition of these residual herbicides improved at harvest control (87 to 96%) when followed by sequential applications of 2,4-D or 2,4-D plus glyphosate; yields from these systems were similar to those with glufosinate plus 2,4-D. Comparison of 2,4-D and 2,4-DB treatments confirmed that 2,4-D is a more effective option for the control of Palmer amaranth. Results from these experiments suggest cotton with resistance to glufosinate, glyphosate, and 2,4-D will improve Palmer amaranth management. At-plant residual herbicides should be recommended for consistent performance of all 2,4-D systems across environments, although cotton with resistance to glyphosate, glufosinate, and 2,4-D will allow greater flexibility in selecting PRE herbicide(s), which should reduce input costs, carryover concerns, and crop injury when compared to current systems.

Eleven field experiments were conducted over a 3-yr period (2010, 2011, and 2012) in conventional- and no-till soybean with a flumioxazin and pyroxasulfone premix. PRE and preplant applications were evaluated for soybean injury, weed control, and yield compared to standard herbicides. Early-season soybean injury from flumioxazin/pyroxasulfone ranged from 1 to 19%; however, by harvest, soybean yields were similar across labeled rates (160 and 200 g ai ha−1), standard treatments, and the nontreated control. Flumioxazin/pyroxasulfone provided excellent control (99 to 100%) of velvetleaf, pigweed species (redroot pigweed and smooth pigweed), and common lambsquarters across almost all rates tested (80 to 480 g ai ha−1). Common ragweed, green foxtail, and giant foxtail control increased with flumioxazin/pyroxasulfone rate. The biologically effective rates varied between tillage systems. The flumioxazin/pyroxasulfone rate required to provide 80% control (R80) of pigweed was 3 and 273 g ai ha−1 under conventional- and no-till, respectively. For common ragweed, the R80 was 158 g ai ha−1 under conventional tillage; yet, under no-till, the rate was nonestimable. The results indicate that flumioxazin/pyroxasulfone can provide effective weed control as a setup for subsequent herbicide applications.

Johnsongrass is a troublesome weed infesting spring–summer crops. Poor control of johnsongrass after fluazifop-p-butyl treatments has been reported in central to northern Italy. Greenhouse and outdoor dose–response experiments revealed that four populations were highly resistant to fluazifop-p-butyl. All four were cross-resistant to other aryloxyphenoxypropionate (FOP) herbicides—propaquizafop, quizalofop, and haloxyfop. The resistance indexes ranged between 8 and 25 for propaquizafop and quizalofop, whereas a greater variability between populations was detected in response to haloxyfop. Conversely, cycloxydim and clethodim determined only a shift in the susceptibility with resistance index (RI) values of 2 to 3. Molecular analyses revealed that resistant plants possessed an insensitive acetyl coenzyme-A carboxylase (ACCase) target enzyme due to an Ile-to-Asn substitution at codon 2041. To our knowledge, this is the first report of such a mutation endowing ACCase resistance in johnsongrass. A molecular marker (CAPS assay) was developed for its rapid detection. Alternative mode of action herbicides S-metolachlor and nicosulfuron controlled all the FOP-resistant populations. Only a few chemical options are still available, and they have different efficacy on germinating seeds and sprouting rhizomes. To maintain efficacy over time, herbicides should be integrated with agronomic practices.

Glyphosate-resistant Palmer amaranth escaping residual herbicides is difficult to manage in cotton because of its rapid growth and a limited number of effective herbicide options to control emerged plants. An experiment was conducted at two dryland and two irrigated sites in Georgia during 2011 and 2012 to determine if cotton resistant to glyphosate, 2,4-D, and glufosinate could be used to salvage a crop infested with large Palmer amaranth. Three POST herbicide systems, including sequential applications of 2,4-D, sequential applications of 2,4-D plus glufosinate, or 2,4-D followed by (fb) glufosinate, were applied with intervals of 5, 10, or 15 d between POST applications. All three systems were followed by diuron plus MSMA directed at layby. At the dryland sites with high temperatures and drought conditions, no program provided greater than 90% control. However, the 2,4-D plus glufosinate system was at least twice as effective in controlling 20-cm-tall Palmer amaranth and produced at least three times more cotton than the other two systems, when pooled over POST application intervals. Intervals of 10 or 15 d between POST applications were 23 to 27% more effective than a 5-d interval in controlling Palmer amaranth when pooled over POST herbicide systems; yields were nearly twice as much with the 10-d interval compared to 5 d. At the irrigated site, overall weed control was greater with less treatment differences noted. Palmer amaranth that was 20 cm tall at application was controlled 98 to 99%, 92 to 93%, and 81 to 94% by glufosinate plus 2,4-D, 2,4-D fb glufosinate, and 2,4-D fb 2,4-D systems at harvest, respectively. Intervals between POST applications only influenced control by the POST 2,4-D system, and the 10-d interval was more effective than the 5-d interval. Carpetweed, Florida beggarweed, and smallflower morningglory were controlled 99% at harvest by all systems; however, it was noted that control of carpetweed and Florida beggarweed prior to layby was less effective with 2,4-D than systems including glufosinate. In the event of an at-plant residual herbicide failure in fields infested with glyphosate-resistant Palmer amaranth, our research demonstrates that glufosinate plus 2,4-D sequentially applied 10 to 15 d apart followed by a timely layby application controlled the target weeds in cotton with resistance to 2,4-D, glyphosate, and glufosinate.

Field experiments were conducted in 2008, 2011, and 2012 to investigate the interaction between seeding system disturbance and PPI herbicides on rigid ryegrass control in wheat. Of the herbicides examined, prosulfocarb + S-metolachlor, and pyroxasulfone provided ≥ 70% control of rigid ryegrass, irrespective of seeding system. In contrast, trifluralin was the least-effective herbicide against rigid ryegrass and was particularly ineffective when used with single disc (10% control) relative to the triple-disc seeding system (80%) in 2012. Trifluralin consistently reduced wheat density when incorporated using single discs (46 to 59%) but not do so with the triple disc or double-shoot knife-point systems. Although there were large differences in crop establishment because of herbicide phytotoxicity, that did not always translate into large differences in yield because wheat was able to recover from reductions in plant density by increasing the spike number per plant. Pyroxasulfone caused no damage to wheat and appeared to be the most suitable PPI herbicide for use with single-disc seeding systems.

In Australia, most wheat is sown in a no-till system without prior cultivation where herbicides are applied prior to sowing and incorporated by the planter. Trifluralin has been the most widely used PRE herbicide to control rigid ryegrass. The objective of this research was to determine crop safety and efficacy of alternative mechanism of action PRE herbicides for rigid ryegrass control in no-till wheat production. Pyroxasulfone achieved 98% control with PRE applications. The alternative PRE herbicides tested alone and in mixtures occasionally resulted in a significant reduction in wheat emergence but not crop yield. Trifluralin treatments failed at sites having trifluralin-resistant rigid ryegrass. Pyroxasulfone and prosulfocarb plus S-metolachlor were effective for control of rigid ryegrass across all trials with control ranging from 64 to 94%. This research demonstrated that PRE applications of herbicides other than trifluralin such as pyroxasulfone and prosulfocarb plus S-metolachlor can be safely and effectively used to control rigid ryegrass in no-till wheat.

Flumioxazin provides PRE and POST, annual weed control in dormant bermudagrass, but applications during active growth may be injurious. Flumioxazin could also provide an alternative chemistry for POST annual bluegrass control in other turfgrasses, but research is limited on tolerance levels. The objective of this research was to evaluate tolerance of five warm-season turfgrasses to flumioxazin applied at various rates and timings. Late-winter applications of flumioxazin at 0.21, 0.42, or 0.84 kg ai ha−1 caused acceptable (< 20%) injury to bermudagrass, seashore paspalum, St. Augustinegrass, and zoysiagrass at 3, 6, and 9 wk after treatment (WAT) in both years. In 2012, late-winter applications to centipedegrass caused unacceptable injury at 6 WAT, but turf recovered to acceptable levels by 9 WAT at all rates. Applications made during active turfgrass growth caused unacceptable initial injury to all species. However, bermudagrass, seashore paspalum, St. Augustinegrass, and zoysiagrass recovered to < 20% injury by 9 WAT from all rates. In 2012, centipedegrass treated in midspring had 0, 24, and 74% injury from flumioxazin at 0.21, 0.42, and 0.84 kg ha−1, respectively, at 9 WAT. In 2013, midspring applications to centipedegrass caused 13, 48, and 71% injury from 0.21, 0.42, and 0.84 kg ha−1, respectively at 9 WAT. Overall, flumioxazin has the potential to control annual weeds in bermudagrass, seashore paspalum, St. Augustinegrass, and zoysiagrass with late-winter applications before greenup, but all turfgrasses may be excessively injured during active growth.

Field trials were initiated in fall 2011 to determine the potential of pyroxasulfone to control acetolactate synthase (ALS) inhibitor-resistant weeds in field pea. Pyroxasulfone was applied in split-plot trials at five locations in western Canada using fall and PRE spring applications of 0 to 400 g ai ha−1. Trial locations were chosen with a range of soil organic matter content: 2.9, 4.3, 5.5, 10.5, and 10.6% at Scott, Kernen, Kinsella, Melfort, and Ellerslie, respectively. The herbicide dose required to reduce biomass by 50% (ED50) in false cleavers ranged between 53 and 395 g ha−1 at Scott and Ellerslie, respectively. Wild oat ED50s varied between 0.54 g ha−1 at Scott in the fall and 410 g ai ha−1 in the spring at Melfort. ED50s for wild oat and false cleavers varied by 7.4- and 746-fold, respectively, depending primarily on the organic matter content at the trial location. The effect of application timing was not consistent. Significant yield reductions and pea injury occurred at 150 and 100 g ha−1 and higher at Kernen and Scott, respectively. Low organic matter and high precipitation levels at these locations indicates increased herbicide activity under these conditions. Pyroxasulfone may allow control of ALS inhibitor-resistant false cleavers and wild oat; however, locations with high soil organic matter will require higher rates than those with low organic matter for similar control levels.

Dense populations of glyphosate-resistant (GR) Italian ryegrass are problematic for spring burndown herbicide programs and crop establishment in the midsouthern United States. Two field studies were conducted to evaluate fall-applied residual herbicides for control of GR Italian ryegrass and to identify the most effective application timing for these herbicides. Fall applications of clomazone at 0.84 and 1.12 kg ai ha−1, pyroxasulfone at 0.16 kg ai ha−1, and S-metolachlor at 1.79 kg ai ha−1 controlled GR Italian ryegrass ≥ 93% 180 d after application. Control from incorporated applications of pendimethalin at 1.59 kg ai ha−1 and trifluralin at 1.68 kg ai ha−1 and surface applications of S-metolachlor at 1.42 kg ha−1 provided control similar to the best treatments. Glyphosate-resistant Italian ryegrass control following clomazone, pyroxasulfone, S-metolachlor, or trifluralin applied in mid September, October, or November exceeded that from fall tillage by 19 to 56% at 90 and 140 d after the last treatment. Pyroxasulfone and S-metolachlor controlled more GR Italian ryegrass following October or November applications compared with those in September at both 90 and 140 d after the last application timing. However, the benefit of delaying clomazone application from October to November was not realized until the last evaluation (140 d after the last treatment). Clomazone, pyroxasulfone, and S-metolachlor offer growers the best opportunity for residual control of GR Italian ryegrass, and control is optimized when these herbicides are applied in November.

Teff is a warm-season C4 annual grass crop grown for forage and food grain that has recently increased in production in parts of the United States. Hay from teff is well suited for livestock, especially horses. The objective of this study was to evaluate teff and weed response to selected herbicides in field studies conducted at the Malheur Experiment Station, Ontario, OR in 2009 and 2010. Herbicides were applied POST when teff was at the four-leaf stage. Broadleaf weed control at 21 d after treatment was greater than 91% across herbicide treatments. Only the premix of 2.5 g ai ha−1 florasulam + 99 g ae ha−1 fluroxypyr + 15 g ai ha−1 pyroxsulam provided acceptable control of barnyardgrass. Due primarily to barnyardgrass competition, teff treated with a premix of 2.5 g ha−1 florasulam + 99 g ha−1 fluroxypyr + 15 g ha−1 pyroxsulam produced 7,200 kg ha−1 of teff hay compared with 4,800 kg ha−1 of teff hay for 2,4-D and dicamba and 4,200 kg ha−1 teff hay when no herbicides were used. Teff grain production was greater with 2.5 g ha−1 florasulam + 99 g ha−1 fluroxypyr + 15 g ha−1 pyroxsulam compared with any of the other treatments. The use of a premix of florasulam + fluroxypyr + pyroxsulam would improve broadleaf and grass weed control in ‘Tiffany' and ‘Dessie' teff varieties, improve hay and grain yield, and reduce production costs.

Isothiocyanates (ITCs) were evaluated as an alternative to methyl bromide (MeBr) for control of Palmer amaranth, large crabgrass, and yellow nutsedge; reduction of tuber density; and increase in marketable tomato yield in low density polyethylene (LDPE)-mulched tomato production. Allyl ITC was applied at 450, 600, and 750 kg ai ha−1; metham sodium (methyl ITC generator) was applied at 180, 270, and 360 kg ai ha−1; and MeBr plus chloropicrin (mixture of MeBr and chloropicrin at 67 : 33%, respectively) was applied at 390 kg ai ha−1. A nontreated weedy check was included for comparison. There was no injury to tomato plants following allyl ITC, metham sodium, or MeBr application. Allyl ITC at 750 kg ha−1 or metham sodium at 360 kg ha−1 controlled Palmer amaranth ≥ 79%, large crabgrass ≥ 76%, and yellow nutsedge ≥ 80% and was comparable to the weed control with MeBr. Highest rates of allyl ITC and metham sodium reduced yellow nutsedge tuber density (≤ 76 tubers m−2) comparable to the MeBr application. Total marketable tomato yield was ≥ 31.6 t ha−1 in plots treated with allyl ITC at 750 kg ha−1 or metham sodium at 360 kg ha−1. Marketable tomato yield from the highest rate of allyl ITC or metham sodium were similar to the yield (38.2 t ha−1) with MeBr treatment. Therefore, allyl ITC at 750 kg ha−1 and metham sodium at 360 kg ha−1 are effective alternatives to MeBr for Palmer amaranth, large crabgrass, and yellow nutsedge control in LDPE-mulched tomato.

POST control of grass weed species and contaminant turfgrass species, such as common and hybrid bermudagrass, is a major challenge in zoysiagrass. Fluazifop-P-butyl is a POST herbicide that can provide selective grass weed control with limited injury to zoysiagrass. However, because of the extent of genetic variation among zoysiagrass cultivars, it was proposed that fluazifop-P-butyl tolerance was cultivar dependent. Dose–response field studies were conducted in Jay, FL, in 2012 and 2013, evaluating the tolerance of 11 commercial zoysiagrass cultivars and common bermudagrass to 0, 44, 88, 131, 175, and 263 g ai ha−1 of fluazifop-P-butyl. ‘Empire' and ‘Meyer' zoysiagrass were injured 30% or less at 2 wk after treatment (WAT) when treated with 131 g ha−1 of fluazifop-P-butyl or less. Conversely, ‘Zeon' and ‘PristineFlora' reached from 30 to more than 50% injury between 2 and 4 WAT with 44 and 88 g ha−1 fluazifop-P-butyl. ‘Geo', ‘Crowne', ‘JaMur', and ‘UltimateFlora' showed intermediate injury (i.e. ≤ 60% with 175 and 264 g ha−1) and a recovery period longer than 4 WAT. ‘Emerald', ‘Palisades', and ‘Royal' also showed intermediate injury (i.e., 30 to 65% with 175 and 264 g ha−1), which peaked at 2 WAT and then rapidly declined below 25% injury at 4 WAT. By 8 WAT, all zoysiagrass cultivars had recovered, and no injury was observed. Results from a greenhouse experiment demonstrated that differences observed in the field were cultivar related, especially for the lower tolerance level observed in Zeon and PristineFlora. There is potential to increase fluazifop-P-butyl tolerance in zoysiagrass because of the observed variation among 80 zoysiagrass breeding lines.

Western ragweed is problematic in Oklahoma pasture and rangeland. There are no labeled PRE herbicides for western ragweed control in pastures to allow ranchers the flexibility of extending the spray season or the ability to spray before emergence of sensitive crops in bordering fields. Field studies were conducted in four locations during a 2-yr period to determine the effects of applying aminopyralid PRE for control of western ragweed. Good to excellent control of western ragweed was obtained when aminopyralid rates of 89 g ae ha−1 or greater were applied PRE. Control was season long and equal to that obtained with a formulated mixture of 2,4-D and aminopyralid applied POST at a rate of 457 + 57 g ae ha−1. Sward height at time of PRE application had little to no effect on the efficacy of the treatment, but time after application before a significant rainfall event did influence the aminopyralid rate needed to achieve acceptable PRE control.

Water smartweed, an uncommon but troublesome perennial weed of the knotweed family (Polygonaceae), has been endemic with its terrestrial growth form in the area of Kato Nevrokopi in northern Greece, one of the most important potato-growing areas of the country. Two field trials were conducted in 2011 to study the response of water smartweed, wild buckwheat (an annual weed also of the knotweed family), and two potato cultivars (‘Agria' and ‘Banba') to various mixtures of metribuzin and pendimethalin, the two most commonly used herbicides for PRE weed management in potato in Greece. Herbicide mixtures consisted of pendimethalin plus metribuzin, pendimethalin plus rimsulfuron, metribuzin plus rimsulfuron, metribuzin plus prosulfocarb, metribuzin plus flufenacet, and pendimethalin plus metribuzin plus rimsulfuron applied preemergence. Metribuzin and pendimethalin were also applied alone. Water smartweed was difficult to control at the treatments and rates tested. Among all mixtures, a premixture of metribuzin plus prosulfocarb at 320 + 3,200 g ai ha−1 provided the greatest control of water smartweed (47%) and wild buckwheat (87%). Herbicide treatments did not cause detrimental effect on growth of either potato cultivar, and marketable tuber yield generally improved with application of mixtures when compared with metribuzin or pendimethalin applied alone. Control of water smartweed in northern Greece potato production remains a difficult problem to solve, suggesting that until new herbicides are registered for use in potato in Greece, other options of weed control should be sought.

A study was conducted to quantify the magnitude and sources of variation in common waterhemp temporal patterns of emergence over 1 yr. In 2008 and 2010, emergence patterns in the absence of soil disturbance were determined for replicated samples of maternal families (progeny from one individual) separately harvested during the previous year from four plants within each of four agricultural fields (16 maternal families yr−1) at a university research farm near Urbana, IL. Combining data across years, variance partitioning indicated that seed sample within maternal family explained 48% of total variation in the percentage of viable, buried seeds that produced seedlings. Differences within, rather than among, maternal families also accounted for large fractions (60 to 99%) of total variation in cumulative percentage emergence at specific points during the growing season. Within years, seed samples characterized by delayed or accelerated emergence patterns did not originate from specific maternal plants. These results indicate that common waterhemp seed populations are without strong maternal plant effects that limit emergence to narrow intervals within the overall emergence period. Thus, results of this study support the use of contemporary approaches for modeling herbicide resistance evolution in common waterhemp, which assume seedling cohorts contain offspring from all individuals occurring within the maternal population.

In the Pacific Northwest, downy brome continues to infest winter wheat, especially in low-rainfall areas where the winter wheat–summer fallow rotation is the dominant production system. In Washington, a study was conducted for 2 yr at two locations in the winter wheat–summer fallow region to determine the influence of four postharvest tillage treatments on vertical seed movement, seedbank depletion, and plant densities of downy brome. The four tillage implements included a disk, sweep plow, harrow, and skew treader. The study also included a no-till treatment for comparison. The sweep plow and disk led to the most vertical movement of downy brome seed compared with the no-till treatment. Approximately 75% of the fall postharvest seed in the no-till treatment was located either on the soil surface or in the 0- to 3-cm depth at both locations. In contrast, 75% of the seed in the disked treatment was located from 0 to 6 cm deep at both locations. The disk and sweep plow both decreased downy brome seed in the soil at the 0- to 3-cm depth compared with the harrow and no-till treatments. There was no difference in downy brome plant densities following postharvest tillage in the summer fallow due to any of the treatments. However, plant densities in the subsequent winter wheat crop were reduced by the disk and sweep plow compared with the no-till and skew-treader treatments. In general, seed densities as affected by the skew treader fell between the disk and the no-till treatments. The use of the sweep plow and the disk should be integrated into a weed management strategy for downy brome in the wheat–fallow region of the Pacific Northwest.

Field studies were conducted in 2011 and 2012 at the Sandhills Research Station near Jackson Springs, NC to determine the influence of weed-free strip width (WFSW) on newly planted ‘Navaho' blackberry plant growth, fruit yield, and fruit quality. Treatments consisted of 0-, 0.3-, 0.6-, 1.2-, 1.8-, and 2.4-m WFSW. Predicted blackberry yield increased from 718 to 1,015 kg ha−1 at WFSW of 0 to 2.4 m. The currently recommended WFSW of 1.2 m resulted in a blackberry yield of 1,013 kg ha−1. Predicted individual blackberry fruit weight displayed a positive linear response to WFSW and increased from 3.1 to 3.6 g fruit−1 at WFSW of 0 to 2.4 m. Soluble solids content (SSC) of dull black blackberry fruit was greatest (15.1 Brix) when WFSW was 0 m. Relative to a WFSW of 0 m, SSC was reduced 2.3 to 3.4% as WFSW increased from 0.3 to 2.4 m, respectively. WFSW did not influence shiny black blackberry fruit SSC, nor titratable acidity, sugar-to-acid ratio, or pH of shiny or dull black blackberry fruit or primocane number, length, and stem caliper.

Enhanced dissipation of atrazine has been documented in many parts of the world where the herbicide has been extensively used. Atrazine is widely used in corn in Illinois, but enhanced dissipation in the field has not been documented. In this study, the dissipation of atrazine was measured in three fields. Two of the fields (URB-1 and URB-2) had not been treated with atrazine for 3 and 4 yr, respectively, whereas the third field (ORR-1) had received annual applications of atrazine for 9 yr. A laboratory assay conducted on soil collected from these fields prior to atrazine application indicated that soil from ORR-1 had enhanced atrazine dissipation (half life [DT50] 1.7 d). Soil from fields URB-1 and URB-2 dissipated atrazine much slower (DT50 was 13 and 11d, respectively). In the field, the rates of atrazine dissipation were similar to those measured in the laboratory. The DT50s were 2.4, 8.4, and 12.6 d for ORR-1, URB-2, and URB-1, respectively. Testing the soils collected from the fields at the last sampling time for atrazine dissipation showed that all of them had enhanced atrazine dissipation, suggesting that the soil microbial populations had adapted to the presence of atrazine in the field.

Scotch broom is a large, nonnative shrub that has invaded forests and grasslands in 27 U.S. states. Without treatment, Scotch broom's persistent seedbank ensures a continuing source of regeneration after soil disturbance. In growth chamber studies, five rates of three synthetic auxin herbicides, aminocyclopyrachlor (AC), aminopyralid (AP), and clopyralid (CP), were compared for PRE control of Scotch broom. Cumulative 90-d emergence, mortality, and biomass of seedlings did not vary among herbicides, averaging 42% of seeds sown, 75% of emerged seedlings, and 9 mg seedling−1 for treated containers, respectively, versus 46%, 17%, and 26 mg seedling−1 for nontreated containers. Low rates of each herbicide (< 50% of the maximum labeled rate [MLR]) provided 60 to 80% control, whereas 100% MLR provided 69 to 89% control. Although the herbicides differed in cost per unit seedling mortality (AP < CP < AC), each demonstrated strong potential as a viable treatment for PRE control of Scotch broom.