A non-crop study was conducted during 2015 and 2017 at the Ponder Research Farm located near Ty Ty, Georgia (31.507654 N, -83.658395 W) on a Tifton loamy sand (fine-loamy, Kaolinitic, thermic Plinthic Kandiudult) with 93% sand 3% silt, 4 % clay, 1% organic matter, and pH of 6.0. The trial was arranged in a randomized complete block design with a 3 (herbicide treatments) by 4 (different times during the d) factorial arrangement of treatments. The herbicide treatments included: paraquat (0.21 kg ai/ha) plus bentazon (0.37 kg ai/ha) plus acifluorfen (0.19 kg ai/ha) plus S-metolachlor (1.23 kg ai/ha); imazapic (0.07 kg ai/ha) plus S-metolachlor (1.23 kg ai/ha) plus 2,4-DB (0.25 kg ai/ha); and lactofen (0.22 kg ai/ha) plus S-metolachlor (1.23 kg ai/ha) plus 2,4-DB (0.25 kg ai/ha). A non-treated control (NTC) was included for comparison. Application timing were at 7:00, 12:00, 17:00, and 22:00 hr. Temperature, relative humidity (RH), and weed height for each application are presented in Table 1.

Plot size was 7.6 m by 0.9 m. Each treatment was replicated 3 or 4 times depending upon field size. Palmer amaranth and a non-uniform mixture of annual grasses including; Texas millet (Brachiaria texana, Buckley), crowfootgrass (Dactyloctenium aegyptium, L. Wild), goosegrass (Eleusine indica, L. Gaertn.), and crabgrass (Digitaria spp.) were present in the non-treated check plots at densities of 50 - 100 plants/m² and 20 - 40 plants/m², respectively. The treatments were applied using a CO₂-pressurized backpack sprayer calibrated to deliver 141 L/ha at 262 kPa and 4.83 km/ha. An 11002DG flat fan nozzle was used for all applications (TeeJet, Springfield, IL 62701). Visual estimates of percent weed control were obtained at 7 and 14 d after treatment (DAT) using a scale of 0% = no control; 100% = complete control or plant death. Plant stunting, chlorosis, and necrosis were considered when making the visual estimates.

An in-crop trial was also conducted at the Ponder Research Farm and the Attapulgus Research and Education Center (30.763629⁰N, -84.479938⁰W) on a Faceville loamy sand (Fine, kaolinitic, thermic Typic Kandiudults) with 84% sand, 10% clay, 6% silt, 1.6% organic matter, and pH of 6.0 during 2015, 2016, and only at the Ponder Research Farm in 2017 (4 site-years). Conventional tillage practices were used and 'Georgia-06G' (Branch, 2007) peanut was planted at both locations. A vacuum planter (Monosem Precision Planters, 1001 Blake St., Edwardsville, KS 66111) was calibrated to deliver 18 peanut seed/m at a depth of 5 cm. Peanut was planted in 2 twin rows (90 cm by 22 cm spacing) at Ponder and 2 single rows (90 cm spacing) in Attapulgus. Plot size was 7.6 m by 0.9 m.

The trial was arranged in a randomized complete block design with a 2 (herbicide programs) by 4 (different TOD) factorial design with 4 replications. The herbicide programs used are presented in Table 2. Each herbicide program was applied at each TOD throughout the entire season (7:00, 12:00, 17:00, and 22:00 hr). Temperature, RH, and weed size at each treatment are presented in Table 3. Herbicides were applied using a CO₂-pressurized backpack sprayer calibrated to deliver 141 L/ ha at 262 kPa and 4.83 km/h_. Visual estimates of peanut crop injury were obtained 7 to 14 after the EPOST and 7 to 14 d after POST treatments. Visual estimates of crop injury consisted of a combination of leaf burn and stunting (0%= no crop injury; 100%= no crop present). Visual estimates of weed control were recorded at 7 and 14 d after the EPOST treatment and 7 to 14 and 50 d after the POST treatment. Weed species were rated including Palmer amaranth, sicklepod, and a non-uniform mixture of annual grasses including, Texas millet, crowfootgrass, goosegrass, and crabgrass. Peanuts were inverted, allowed to air dry, and harvested 4 d later using commercial equipment. Peanut yields were adjusted to 10% moisture.

University of Georgia Extension peanut production recommendations were used and supplemental irrigation was applied to maximize peanut growth and development (Anonymous 2017). Soil characteristics, planting date, peanut stages of growth at application, weed heights, and harvest dates are presented in Table 4 (Boote 1982).

Data for all parameters in both the non-crop and in-crop studies were analyzed as factorial plot designs and subjected to ANOVA using the PROC MIXED procedure in SAS (SAS Institute 107 Inc., Cary, NC 27511). TOD and herbicide treatment/program were considered fixed effects and locations and replications (nested within year) were considered random effects. Least square means of significant main effects were separated using Tukey's HSD test (alpha=0.10).

For the non-crop study there was a significant interaction between TOD and herbicide treatment for the 7 DAT rating for Palmer amaranth control. Palmer amaranth control was lower when imazapic plus S-metolachlor plus 2,4-DB were applied at 22:00 hr, when compared to applications made during daylight hours. Palmer amaranth control was unacceptable (< 70%) with imazapic because the population at this location is known to be ALS-resistant (Wise et al. 2009). At 14 DAT, there was no interaction between treatment and TOD. Palmer amaranth control was reduced with the combination of imazapic plus S-metolachlor plus 2,4-DB when compared with other treatments and no TOD effects were observed. Previous research has shown that reduced control of barnyardgrass can occur when using the ALS-herbicide nicosulfuron at different TOD (Stewart et al., 2009). There is no published evidence that barnyardgrass exhibits diurnal leaf movements, therefore reduced herbicide interception is not the cause of the reduction in control (Mohr et al., 2007).

There was no significant interaction between TOD and treatment for the 7 and 14 DAT rating timing for annual grass control. At 7 DAT, annual grass control was reduced with the imazapic plus S-metolachlor plus 2,4-DB and lactofen plus s-metolachlor plus 2,4-DB treatments. At 14 DAT, only the lactofen plus S-metolachlor plus 2,4-DB treatment provided unacceptable control of annual grasses (< 35%). Lactofen is a broadleaf herbicide and has limited efficacy on grass weed species (Grichar, 1991; Minton et al., 1989). TOD had no effect on annual grass control. This is contrary to previous research where a TOD effect was observed for barnyardgrass control with nicosulfuron (Stewart et al., 2009).

Peanut crop injury was evaluated 1 wk after the EPOST and POST applications. Significant differences in injury were observed at both times. Generally, herbicide programs were less injurious when applied at 7:00 hr and 22:00 hr (Tables 5 and 7). When averaged over TOD, the lactofen program was more injurious than the imazapic program (Table 6). Peanut injury from lactofen has been observed in other research ranging from 20 to 48%, with no observed yield losses (Boyer et al., 2011; Ferrell et al., 2013). However, yield losses of 5% were observed from applications of lactofen applied approximately 60 d after planting (Dotray et al., 2012).

Palmer amaranth was completely controlled by a combination of the PRE and EPOST herbicide applications both 1 and 2 wk after the EPOST application was made (data not shown). For Palmer amaranth control after the POST herbicide applications, there was no significant interaction for any rating. There were also no significant differences between programs or TOD (Tables 6 and 7). Reduced control of Palmer amaranth when applying 2,4-D, imazethapyr, dicamba, glufosinate, and bentazon late in the evening or at night, has been observed in other research (Doran and Andersen, 1976; Johnston et al., 2018; Montgomery et al., 2017; Stopps et al., 2013). However, these studies differ from ours in that a single herbicide was used, not a tank-mixture with multiple active ingredients. TOD does not appear to influence Palmer amaranth control when a complete peanut herbicide program is used consisting of multiple active ingredients at a single application timing.

After the EPOST applications, annual grass control at 7 to 14 DAT was significantly lower at 7:00 hr (Table 5). Paraquat has been reported to be more efficacious when applied at night or later in the evening for weed control due to minor intercellular translocation occurring (Brian, 1967; Montgomery et al., 2017; Putnam and Ries, 1968). It has been reported that annual grasses are not as sensitive as broadleaf weeds to TOD effects with herbicides (Stewart et al., 2009). One possible explanation for the reduction in control observed at 7:00 hr is that dew was present on the weeds and on the crop. Dew presence has been reported to both increase or decrease herbicide efficacy depending on the herbicide and weed species involved (Caseley, 1989; Fausey and Renner, 2001; Nalewaja et al., 1975; Wanamarta and Penner, 1989). There was no significant TOD by program interaction for grass control after the POST applications, thus data is averaged over TOD and program. At 1 wk after the POST application there was no difference in herbicide program for annual grass control (Table 6). However, at the end of season rating (50 DAT) there was a significant difference in control of annual grasses. Annual grass control with the lactofen program was significantly lower than the imazapic program. Although primarily used for nutsedge (Cyperus spp.) and broadleaf weed control in peanut, imazapic provides various levels of annual grass control depending upon the species and stage of growth (Monks et al., 1996; Wilcut et al., 1999; Jordan et al., 2009). Lactofen has also been reported to have limited activity on grasses (Grichar, 1991; Minton et al., 1989).

At 7 to 14 d after the POST application, reduced control of annual grass was observed at 7:00 and 22:00 hr (Table 7). However, at the end of season control rating, there was no significant difference in TOD for annual grass control. As previously mentioned grass species are less sensitive to TOD effects when compared to broadleaf weed species (Stewart et al., 2009). Additionally, the dense canopy and spreading growth habit of the peanut crop likely resulted in the peanut plants out-competing the few grasses that were present at the earlier application timing (Leon et al., 2016).

There was a significant TOD effect for control of sicklepod at 7 to 14 d after the EPOST application with less control, observed at 22:00 hr (Table 5). The diurnal leaf movement of sicklepod has been reported to reduce herbicide interception and control (Norsworthy et al. 1999). There was no interaction between herbicide program and TOD for the visual control ratings of sicklepod after the POST application was made, data presented are pooled over herbicide treatment and TOD. Differences for sicklepod control were observed for both program and TOD at 7 to14 and 50 d following the POST application. The application made at 22:00 hr resulted in less sicklepod control when compared to all other application timings. The imazapic program was more effective than the lactofen program for the control of sicklepod (Table 6). Sicklepod control with imazapic has been well documented (Grey et al., 2003; Grey and Wehjte, 2005).

There was no interaction between TOD and herbicide program for peanut yield; data are pooled over TOD, herbicide program, and site-years. Herbicide program had a significant effect on peanut yield. The imazapic program resulted in higher yields than the lactofen program. The reduction in yield observed between the two programs is potentially due to less annual grass and sicklepod control that was observed between the two treatments. Control of annual grass species for 8 to 10 wks after peanut emergence has been shown to be critical in maintaining a high yielding peanut crop (Everman et al., 2008; Grichar, 1991; Johnson and Mullinix, 2006). Peanut injury from lactofen has been observed in other research ranging from 20 to 48%, with no observed yield losses (Boyer et al. 2011; Ferrell et al. 2013). But other research has reported a yield loss of 5% from lactofen 60 DAP (Dotray et al. 2012). TOD of did not affect peanut yield. While reductions in sicklepod control were observed from applications made at 22:00 hr those reductions did not result in yield loss.