Field studies were conducted during the 2003 through 2005 growing seasons to evaluate postemergence herbicides and timing of application for horse purslane, smellmelon, and Palmer amaranth control in peanut. Acifluorfen provided inconsistent control of the three weed species. Bentazon failed to control Palmer amaranth, horse purslane, and smellmelon (< 65%), whereas 2,4-DB controlled horse purslane and smellmelon less than 70%. Diclosulam applied early postemergence (EPOST) controlled Palmer amaranth at least 77% but horse purslane control varied from 27 to 73% when applied EPOST and was less than 50% with the late postemergence (LPOST) application. Diclosulam controlled smellmelon less than 65%. Imazethapyr and imazapic controlled Palmer amaranth and smellmelon at least 70% but controlled horse purslane less than 65%. Lactofen controlled Palmer amaranth and horse purslane at least 93% with the EPOST application but had inconsistent control (40 to 100%) when applied LPOST. Smellmelon control with lactofen was also inconsistent. Removal of weed interference with all herbicides except bentazon provided peanut yields greater than the nontreated check.

Bird vetch is a perennial Eurasian plant which, unlike many exotic weed species, can invade low fertility areas that have not been disturbed. It also is common in pastures, woodland, and tall forb communities. Bird vetch is expanding along Alaskan roadsides, in urbanized areas, and in low density forests. A greenhouse study was conducted to determine efficacy of six herbicide treatments applied at reduced rates in 2005 and again in 2006 for bird vetch seedling control. Bird vetch seedlings were tolerant of reduced rates of chlorsulfuron and 2,4-DB; however, complete control was achieved with rates of clopyralid, dicamba plus diflufenzopyr, triclopyr, and 2,4-D that were a fourth to an eighth of the full registered rate. These results will be important for developing effective, low-cost methods for controlling bird vetch in Alaska, especially on the outer margins of infestations.

Field studies were conducted near Clayton, Goldsboro, Kinston, and Rocky Mount, NC in 2003 to evaluate weed control and cotton response to postemergence (POST) treatments of glufosinate applied alone or in tank mixtures with s-metolachlor, pyrithiobac, or trifloxysulfuron. Late-season control of common lambsquarters, common ragweed, entireleaf morningglory, ivyleaf morningglory, jimsonweed, pitted morningglory, purple nutsedge, and sicklepod with glufosinate early postemergence (EPOST) was ≥90%. The addition of S-metolachlor to glufosinate EPOST improved control of all weeds except sicklepod, ivyleaf morningglory, and entireleaf morningglory. When applied POST, glufosinate provided ≥90% late season control of common lambsquarters, common ragweed, entireleaf morningglory, ivyleaf morningglory, jimsonweed, large crabgrass, pitted morningglory, purple nutsedge, and sicklepod. Control of goosegrass and Palmer amaranth was 81 and 84%, respectively. When pyrithiobac or trifloxysulfuron were added in POST tank mixtures, control of Palmer amaranth improved 6 and 9 percentage points, respectively. Control of goosegrass remained near 80% regardless of herbicide treatment used. The addition of a late post-directed (LAYBY) tank-mixture of glufosinate plus prometryn provided ≥88% late season control of all weeds. Reduced control of goosegrass and Palmer amaranth was observed with the LAYBY tank mixture of glufosinate plus MSMA when compared to other LAYBY tank mixtures. Cotton lint yields in plots receiving any herbicide application were significantly higher than plots receiving no herbicide application for all application timings. Cotton lint yields were ≥ 740 kg/ha where an EPOST was applied and ≥ 680 kg/ha when a POST herbicide was applied. Cotton lint yields were at least 200 kg/ha greater on plots receiving a LAYBY application when compared to plots where no LAYBY treatment was applied.

Annual bluegrass is a troublesome weed of Kentucky bluegrass because of a lack of selective POST herbicides for control. Field experiments were conducted in New Jersey to investigate the potential of bispyribac-sodium, primisulfuron, and sulfosulfuron for selective annual bluegrass control in Kentucky bluegrass. Primisulfuron provided the best combination of Kentucky bluegrass safety and annual bluegrass control with greater efficacy in summer than fall. Bispyribac-sodium also provided substantial annual bluegrass control, especially in the summer, but caused unacceptable (> 20%) Kentucky bluegrass injury, whereas annual bluegrass control with sulfosulfuron was inconsistent. Growth chamber experiments confirmed that greater primisulfuron efficacy in summer than fall may be attributed to higher temperatures. Overall, primisulfuron has promising implications for future use in Kentucky bluegrass for annual bluegrass control whereas bispyribac-sodium will likely be more applicable in other cool-season turfgrasses.

In 2004 and 2005, glyphosate + diuron and flumioxazin + MSMA were applied POST-directed to cotton at 20, 30, 40, and 50 cm in height. Herbicides were directed to either the lower 2 cm of the cotton stem or to the lower 50% of the cotton stem to determine the impact of application timing and placement on cotton response. Glyphosate + diuron resulted ≤ 22% injury when applied high on the stem and ≤ 8% injury when applied low on the stem. Regardless of application placement or timing, no yield reductions were observed as a result of glyphosate + diuron application. Conversely, flumioxazin + MSMA applications resulted in 93 and 64% injury when applied high POST-directed (HPD) to 20- and 30-cm cotton, respectively. Low POST-directed (LPD) applications caused 33 and 23% injury to 20- and 30-cm cotton, respectively. In both years, HPD applications of flumioxazin + MSMA on 20-cm cotton resulted in excessive yield loss (94 to 87%), whereas yield loss was only noted in 30-cm cotton in 2004. No yield loss was observed in 40- or 50-cm cotton. Additionally, LPD applications of flumioxazin + MSMA, though causing significant visual injury, did not result in a yield reduction at any application timing. Although cotton became more tolerant to flumioxazin + MSMA as the season progressed, the research suggests that applications of flumioxazin should be made with precision or delayed until cotton reaches 40 cm in height.

Only two spinach herbicides are currently available to producers, and with limited choices in herbicide mode of action (MOA), overuse of the current herbicides can result in dominance of uncontrolled weeds. Therefore, spinach producers need more herbicide options. Trials were conducted in processing spinach to evaluate the crop safety of dimethenamid-p (0.56 kg ai/ha), EPTC (2.94 to 3.40 kg ai/ha), or ethofumesate (0.84, 1.12 or 2.24 kg ai/ha) applied pre-emergent (PRE), and ethofumesate (0.09 or 0.18 kg ai/ha) applied early postemergent (EPOST) for potential use as alternatives to s-metolachlor (0.73 kg ai/ha) and cycloate (2.52 kg ai/ha). The control of London rocket was excellent with s-metolachlor, dimethenamid-p, and ethofumesate. Weed control with cycloate and EPTC was fair, and generally inferior to all others. Cycloate and s-metolachlor were safe to spinach, and only minor crop injury was observed with all other herbicides, except the high rate of ethofumesate PRE, which resulted in high levels of injury. Spinach outgrew all early herbicide injury with no significant yield losses. Ethofumesate applied EPOST caused no injury in year 1, although in year 2 injury was higher in all plots where it was applied twice, regardless of rate. In high-density plantings, dimethenamid-p caused significant crop injury to spinach, but only at one location on a sandy loam soil. In that trial, yields in dimethenamid-p plots were reduced an average 30% compared to the hand-weeded and s-metolachlor plots. As a result of this research, dimethenamid-p, EPTC, and ethofumesate should be considered candidates for expanded-use registrations in processing spinach production. More research is needed to determine additional rates and use patterns for improved crop safety, as well as an evaluation of soil types and spinach varieties.

Field studies were conducted near Oxnard, CA and in two locations near Salinas, CA in 2002 and 2003 to evaluate efficacy and safety of six herbicides in celery. s-metolachlor at 0.6 and 0.7 kg ai/ha and flufenacet at 0.5, 0.6, and 0.7 kg ai/ha PRE were safe to celery and controlled 81 to 94% of yellow nutsedge at densities < 20 plants/m2. At yellow nutsedge densities > 20 plants/m2, the most efficacious treatments were 1.1 kg/ha of s-metolachlor or 0.7 kg/ha of flufenacet, which reduced nutsedge densities by 71 or 53%, respectively; however, both resulted in slight injury to celery. All other herbicides generally did not injure celery and none reduced marketable crop yield. Flumioxazin pretransplant (PRE) at 0.1 and 0.2 kg ai/ha controlled broadleaf weeds near 100% at all locations, suggesting that it can be an effective alternative to standard linuron and prometryn, which are applied post-transplant (POST) in celery.

Field trials were initiated at three Missouri locations in the fall of 2004 and 2005 to compare the efficacy of fall and early spring herbicide applications on winter and summer annual weed populations in no-till soybean. All three experiments received applications of chlorimuron plus sulfentrazone plus 2,4-D; chlorimuron plus tribenuron plus 2,4-D; and glyphosate plus 2,4-D in the fall, 60 d prior to planting (60 PP), 30 days prior to planting (30 PP), and 7 d prior to planting (7 PP). Good control of all winter annual weeds, except for common chickweed, was observed at planting from fall applications of chlorimuron plus sulfentrazone plus 2,4-D and chlorimuron plus tribenuron plus 2,4-D. Fall, 60 PP, and 30 PP applications of chlorimuron plus sulfentrazone plus 2,4-D provided relatively poor control of common chickweed. Glyphosate plus 2,4-D applied in the fall and 60 PP provided adequate control of early fall germinating winter annuals, such as common chickweed and henbit. However, these treatments provided poor control of winter annual species that exhibited some degree of spring germination. Control of summer annual weed species 7 wk after planting (WAP) was highly variable. Relatively poor control of common waterhemp was obtained from all treatments except chlorimuron plus sulfentrazone plus 2,4-D applied at all three spring timings. Giant foxtail control was highly variable ranging from 0 to 95%. Measurements of weed biomass collected 7 WAP revealed larger amounts of biomass present in the fall application timing of each herbicide treatment. Emergence of winter and summer annual weed species from early March to 7 WAP varied by herbicide treatment. Emergence of total winter and summer annual weed species was reduced with both residual herbicide treatments when compared with glyphosate plus 2,4-D and the untreated control.

Field and greenhouse studies examined glyphosate and several ACCase-inhibitor herbicides for control of wirestem muhly. A field study evaluated application rate and timing for control in glyphosate-resistant soybean. Herbicides were applied at two rates, 4 and 6 wk after planting (WAP) of soybean when wirestem muhly was about 30 and 50 cm tall, respectively. A complimentary greenhouse experiment with the same herbicides examined wirestem muhly rhizome viability 30 days after herbicide application. A second field study examined optimum glyphosate application timing for control of wirestem muhly during a fallow rotation period. Glyphosate was applied at 14-day intervals from mid-May through early October. In the first field study, glyphosate, fluazifop, and clethodim provided better control of wirestem muhly than sethoxydim and quizalofop. Clethodim, fluazifop, and glyphosate provided 74 to 92% control over both rates and application timings, whereas quizalofop and sethoxydim were the least effective, providing only 52 to 64% control regardless of rate or timing. In the greenhouse, glyphosate was the most effective herbicide, providing at least 97% control and an 87% reduction in shoot biomass; sethoxydim and fluazifop were the least effective herbicides. All of the herbicides reduced rhizome viability compared to nontreated plants up to 71%, but no difference between herbicides was observed. In the second field study, end-of-season ratings revealed that glyphosate provided 95 to 100% control of wirestem muhly from applications made between late June and early September. The year following application, glyphosate provided 91 to 99% control of wirestem muhly from applications made between mid-June and early September. Applications earlier than mid-June or after early September resulted in less than 90% control. Overall, results show that (1) glyphosate was the most effective herbicide; (2) the ACCase-inhibitor herbicides differ in their effectiveness; and (3) rate and application timing can impact control.

Mesotrione is under evaluation for registration in turfgrass for weed control, but often requires repeat treatments. Previous research in agricultural crops indicates tank mixtures with mesotrione improve weed control. Greenhouse trials were conducted in Blacksburg, VA, with 7 mo-old white clover, and field trials in Knoxville, TN, evaluated control of 5-yr old white clover in a stand of common bermudagrass. Data from both field and greenhouse trials indicate mesotrione applied in combination with bromoxynil, carfentrazone, or simazine control white clover better than any of these herbicides applied alone. Combinations of bromoxynil, carfentrazone, and simazine with mesotrione also decreased relative chlorophyll index and white clover fresh weights in agreement with visually estimated control.

Field studies were conducted in 2004 and 2005 to determine the effect of soybean row width and population on weeds, canopy closure, crop yield, and economic return in glyphosate-resistant soybean. Soybean leaf area index (LAI) was greater in 19- and 38-cm, compared with 76-cm rows from 8 to 12 wk after planting in the low, moderate, and high soybean populations. Canopy closure was delayed by 2 wk in the moderate population in 76-cm rows compared with the high population in 19-cm rows. Fewer weeds emerged in 19-cm, compared with 76-cm rows following glyphosate application, and increasing the soybean population within a row width did not influence late-season weed emergence. Weed biomass in the weedy control was greater in the very low soybean population compared with the high soybean population within each row width; however, weed biomass in the weedy control was similar in the high and moderate soybean populations. Soybean yield in the weed-free and 10-cm glyphosate treatment did not differ, and yield was greater in 19-cm rows planted at moderate or high, compared with low populations. There was no difference in weed-free soybean yield at low, moderate, and high populations within 38- and 76-cm rows. Gross margins were usually greater in 19- and 38-cm, compared with 76-cm rows. The gross margin for soybean planted in 19-cm rows was usually greater at moderate or high soybean populations compared with lower populations. In 76-cm rows, the gross margin was greatest at the low and moderate soybean populations. When rainfall or other factors limited soybean yield, increasing the soybean population from approximately 300,000 plants/ha to 445,000 plants/ha in 19-, 38-, and 76-cm rows did not result in quicker canopy closure, reduced weed emergence, or greater soybean yield and gross margins.

Field trials were established in Fayetteville, AR, to determine the effects of glyphosate on biomass and seed production of spurred anoda, entireleaf morningglory, hemp sesbania, and Florida pusley. Inconsistent control of these species by glyphosate allows them to enrich the soil seedbank. Glyphosate rates of 0.42, 0.84, and 1.68 kg ae/ha were applied at 3, 6, and 9 wk after emergence (WAE) and were compared with a nonsprayed control. Spurred anoda seed production was reduced 99.9% by glyphosate at 1.68 kg/ha averaged over application timings and 99.7% by glyphosate applied at 6 WAE averaged over rates. Entireleaf morningglory seed production was reduced 100% by glyphosate at 1.68 kg/ha at 3 WAE and > 93% by the same rate applied at 6 or 9 WAE. Hemp sesbania seed production was reduced 100% by glyphosate at 1.68 kg/ha at 3 WAE and > 96% by glyphosate at 0.42, 0.84, and 1.68 kg/ha applied at 6 or 9 WAE. Florida pusley seed production was reduced 99.9% by glyphosate at 1.68 kg/ha averaged over application timings and 99.4% by application at 3 or 6 WAE averaged over rates. Biomass accumulation for all species was reduced most by early applications, particularly at 1.68 kg/ha.

Dallisgrass response to various MSMA and foramsulfuron herbicide combinations was evaluated from 2004 through 2006 in North Carolina. Dallisgrass control declined with herbicide rates; therefore, only the highest rates (foramsulfuron, 0.15 kg ai/ha; MSMA, 2.5 kg ai/ha) are discussed. Foramsulfuron followed by (fb) foramsulfuron 1 wk after initial treatment (WAIT) provided 60% control of dallisgrass 1 mo after initial treatment (MAIT), but control declined to 40% 3 MAIT. MSMA (2.5 kg ai/ha) applied three times provided 89% dallisgrass control 1 MAIT, but control declined to 71% 3 MAIT. Among 22 MSMA and foramsulfuron sequential application programs, MSMA fb foramsulfuron 2 WAIT fb MSMA 3 WAIT provided the highest control (94%) of dallisgrass 1 MAIT, and control levels remained high (93%) 3 MAIT. This herbicide program provided 85% control 1 yr after initial treatment (YAIT), whereas no other treatment provided greater than 37% control 1 YAIT. Timing of foramsulfuron application following the initial MSMA application affected the control of dallisgrass. Waiting 2 WAIT, rather than 1 WAIT, before applying foramsulfuron increased control of dallisgrass by 20% 1 MAIT and 19% 3 MAIT. However, the increase in control 1 YAIT was only 5% regardless of herbicide rate. The addition of a second application of MSMA following the application of foramsulfuron 2 WAIT of MSMA further increased control by 10% 1 MAIT, 37% 3 MAIT, and 48% 1 YAIT. Results from this study suggest dallisgrass may be controlled with applications of MSMA fb foramsulfuron 2 WAIT fb MSMA 3 WAIT applied during early to midsummer. The use of MSMA was required to achieve dallisgrass control, however, the relatively high level of control achieved 1 YAIT with a program including sequential applications of MSMA and foramsulfuron may help reduce the total amount of arsenical herbicides used to control dallisgrass infestations over time.

Field experiments were conducted to evaluate the influence of three nitrogen fertilizer application timings on corn and giant ragweed dry weight, corn and giant ragweed nitrogen accumulation, and corn grain yield. By the V8 growth stage, corn plants in treatments which included nitrogen at planting accumulated up to 27% more dry weight and 26% more leaf area than corn in sidedress-only treatments that had not yet received any fertilizer nitrogen. Nitrogen fertilizer application timing did not influence early-season giant ragweed dry weight, but did influence N accumulation. Giant ragweed at 0.5 plants/m2 accumulated up to 16 kg N/ha in 40-cm-tall plants by the V8 corn growth stage and up to 104 kg N/ha when allowed to remain in corn until the end of the growing season. At this density of giant ragweed, corn yields were not reduced if removed by the V8 corn growth stage. However, season-long interference reduced corn yields by up to 19% regardless of nitrogen application timing. Early-season giant ragweed growth was less responsive to nitrogen application timing than corn, suggesting that nitrogen management programs should not be altered significantly to influence giant ragweed competitiveness in corn. However, giant ragweed should be removed in a timely manner to minimize nitrogen accumulation in nitrogen-limiting corn production situations.

Path rush can become problematic on golf course fairways, especially on roughs where continuous golf cart or maintenance equipment traffic occurs. Research trials conducted in 2003 and 2004 compared the efficacy of 2,4-D and mixtures containing 2,4-D in single or sequential applications to herbicides without 2,4-D for path rush control in bermudagrass roughs. Sequential applications of 1.1 kg ae/ha 2,4-D or 0.8 + 0.2 + 0.1 kg ae/ha 2,4-D + MCPP + dicamba provided 95% path rush control with no injury to bermudagrass, 15 wk after initial treatment (WAIT). Single applications of 1.1 + 0.3 + 0.1 kg/ha 2,4-D + MCPP + dicamba or 1.1 kg/ha 2,4-D provided at least 81% path rush control, 15 WAIT. Treatments without 2,4-D did not provide acceptable levels of control (34% or less, 15 WAIT). 2,4-D applied alone or in preformulated mixtures provided excellent path rush control compared to treatments without 2,4-D. Sequential applications of these treatments provided greater path rush control than single applications. However, only the sequential application of 0.8 + 0.2 + 0.1 kg/ha 2,4-D + MCPP + dicamba provided greater path rush control than the single application. Results from these studies suggest that 2,4-D is a key herbicide for path rush control.

Experiments were conducted in 2005 and 2006 near Huntley, Montana to evaluate two formulations of mesosulfuron-methyl for the control of wild oat in malt barley. The two formulations differed due to the amount of safener, mefenpyr-diethyl, that each contained. Formulation 1 and 2 contained mesosulfuron-methyl and mefenpyr-diethyl at a 1 : 2 and 1 : 6 ratio, respectively. The two formulations were applied alone at two rates or in combination with other small grain herbicides that are typically used for broadleaf weed control. The low and high rate of formulation 1 contained 7.8 and 10.1 g ai/ha of mesosulfuron-methyl and 15.6 and 20.2 g ai/ha of mefenpyr-diethyl, respectively. The low and high rate of formulation 2 contained 2.5 and 3.1 g ai/ha of mesosulfuron-methyl and 15 and 18.6 g ai/ha of mefenpyr-diethyl, respectively. These treatments were compared to a standard treatment of fenoxaprop and a nontreated control. In 2005, formulation 2 applied alone at the low and high rates caused barley injury of 7 and 11% at 21 days after treatment (DAT) and 4 and 7% at 56 DAT, respectively. Applications of formulation 1 caused barley injury that was 9 and 17 percentage points greater at 21 DAT and 12 and 18 percentage points greater at 56 DAT than barley injury from applications of formulation 2 when these herbicides were applied alone at the low and high rate, respectively, in 2005. In both years, formulation 1 applied alone at either rate injured barley more than fenoxaprop at 21 and 56 DAT. In both years, formulation 1 and 2 applied alone controlled wild oat between 85 and 96% at 21 DAT and between 92 and 98% at 56 DAT. Regardless of rate, wild oat control with either formulation was similar to that provided by fenoxaprop at 56 DAT. In 2005 and 2006, barley yield was equivalent between treatments of formulation 1 or 2 and that produced by barley treated with fenoxaprop. No difference in the percentage of plump kernels occurred between treatments of formulation 2 and fenoxaprop. The percentage of plump kernels was reduced 6 to 7 percentage points with treatments containing formulation 1 compared to treatments containing formulation 2. Overall, formulation 2 was effective for the control of wild oat in malt barley. Low levels of malt barley injury occurred with this treatment; however, barley successfully recovered by harvest, and yield and kernel size were equivalent to barley treated with fenoxaprop.

Field studies were conducted during the 2004 and 2005 growing seasons to evaluate the effects of various herbicides on smellmelon control and soybean yield response. Pendimethalin applied pre-emergence (PRE) alone controlled smellmelon no better than 70%, whereas pendimethalin plus imazethapyr, flumioxazin, or cloransulam applied PRE provided almost complete smellmelon control (99%). Pendimethalin applied PRE followed by acifluorfen applied early postemergence (EPOST) provided similar control. Glyphosate alone applied late postemergence (LPOST) provided complete control (100%), and pendimethalin followed by glyphosate applied EPOST controlled smellmelon at least 95%. Most herbicides increased soybean yield over the nontreated check.

Field experiments were conducted in 2003 and 2004 to evaluate the effects of reduced rates of glufosinate on development and yield of three nonglufosinate-resistant cotton varieties. The varieties evaluated were selected on their relative maturity with PayMaster (PM) 1218, early maturity; FiberMax (FM) 960, medium maturity; and Delta and Pine Land (DP) 555, late maturity. Rates of 47, 23, and 4.7 g ai/ha were applied, representing 10, 5, and 1% of the typical use rate per application of 467 g ai/ha, respectively. As might be expected, when averaged over varieties, the 10% rate showed more injury than the 5%, and the 5% rate caused more visual injury than the 1% rate. Pooled over timing and rate, PM1218 showed more injury (18%) than FM960 (7%), which showed more injury than DP555 (1%) 7 days after application (DAA) at the 10% rate. However, although PM1218 showed the most visual injury, this did not translate into delay in maturity or loss of lint yield. DP555 showed 70 kg ai/ha and 50 kg ai/ha lint yield loss when glufosinate was applied at the 10% rate on the fifth and eighth node stage, respectively. DP555 was delayed in maturity when glufosinate was applied at the 10% rate on the eighth node stage. FM960 showed 30 kg/ha lint yield loss when glufosinate was applied at the 10 and 5% rates at the fifth node stage. Maturity of FM960 was delayed with the 10 and 5% rates applied at the fifth node stage.

A study was established to evaluate herbicides registered for barnyardgrass control in rice for activity on creeping rivergrass. Treatments included 208 g ai/ha cyhalofop early POST (EPOST) followed by (fb) 315 g/ha cyhalofop late POST (LPOST), 22 g ai/ha bispyribac EPOST fb 22 g/ha bispyribac LPOST, 66 g ai/ha fenoxaprop EPOST fb 86 g/ha fenoxaprop LPOST, 70 g ai/ha imazethapyr EPOST fb 70 g/ha imazethapyr LPOST and 50 g ai/ha penoxsulam mid-POST (MPOST). Each POST program was evaluated with and without 448 g ai/ha clomazone PRE. Barnyardgrass control, averaged over days after treatment (DAT), was 88 to 92% with the addition of clomazone PRE. Clomazone applied PRE followed by two applications of cyhalofop controlled creeping rivergrass at 91%. Creeping rivergrass control was less than 75% at 35 to 49 DAT. This response indicates only suppression of creeping rivergrass is achieved with the herbicide programs evaluated in this study.

Field and greenhouse studies were conducted from 1996 to 2000 to determine tropical soda apple fruit size required for mature seed, overwintering survival potential of seeds, and growth and fruit production in a containment area near Stoneville, MS (latitude 33°25′N). Seedling emergence was ≥ 50% from tropical soda apple fruit > 1.8 cm in diameter. Tropical soda apple overwintering emergence was 14, 10, 48, 42, and 13% from seed alone and 86, 83, 48, 41, and 18% for seeds in tropical soda apple fruit overwintering at 100, 0, −5, −10, and −15 cm from the soil surface, respectively. Before frost (early November), average tropical soda apple plant heights were 0.8, 0.6, and 0.7 m; plant diameters were 2.6, 1.7, and 2 m; fruit produced per tropical soda apple plant were 415, 342, 128; and plant weights were 9.9, 9.5, and 4.9 kg at 27 wk after transplanting (WATP) for 1996, 1997, and 1998, respectively. Regeneration from overwintering tropical soda apple roots varied depending on the winter conditions. These results suggest that tropical soda apple plants can survive mild winters.