Effect of dose and adjuvant on uptake of triclopyr and dicamba into Pinus contorta needles

Why this research Matters Management of dense infestations of wilding Pinus contorta in New Zealand requires high doses of herbicides; 18 kg active ingredient (a.i.) ha−1 triclopyr and 5 kg a.i. ha−1 dicamba are used in combination with a complex mix of adjuvants (methylated seed oil, non‐ionic surfactant and ammonium sulfate) and other active ingredients. From the perspective of cost and environmental impact there is a need to reduce the complexity of this tank mix and the rates of active ingredients. Using radiolabelled herbicides, this study evaluated the effect of dose and adjuvants (crop oils, non‐ionic surfactants, and organosilicones) on needle injury and uptake of triclopyr and dicamba into P. contorta needles at 24 hr or 7 days after treatment (DAT). The uptake of triclopyr decreased significantly with increasing concentration (0.75%–6%) resulting in the highest uptake dose at the equivalent of 18 kg a.i. ha−1 triclopyr at 7 DAT. When applied at 18 kg a.i. ha−1, none of the adjuvants tested significantly increased the uptake of triclopyr (applied as Grazon®), with ~50% uptake occurring at 7 DAT. The uptake of dicamba (applied as Kamba® at 5 kg a.i. and 10 kg a.i. ha−1) was significantly affected by dose and adjuvants. The uptake of dicamba applied at 5 kg a.i. ha−1 was low at 7 DAT with no adjuvant (31%); however, use of a methylated seed oil doubled the uptake. When triclopyr and dicamba were applied together, there was no evidence that either active ingredient negatively affected uptake of the other, with triclopyr enhancing uptake of dicamba. These results show potential to reduce the amount of herbicide used for conifer control without compromising efficacy.

decades (Ledgard, 2008;Ministry for Primary Industries, 2015), especially as they pose a significant threat to biodiversity, conservation and grazing potential on high country farms. The costs to stop and reduce the spread of these exotic conifers are, and will continue to be, high. However, there will be significant negative impacts on a range of social, economic, and environmental outcomes if spread continues. The need to effectively manage the spread of exotic conifers is not restricted to New Zealand, with South Africa, Sweden, Chile, and Brazil facing similar challenges with respect to the containment of invasive tree weeds (Engelmark et al., 2001;Ledgard, 2001;Rouget, Richardson, Milton, & Polakow, 2004;Simberloff et al., 2013).
Both triclopyr and dicamba are selective, systemic post-emergence, synthetic auxin herbicides that kill plants by inducing abnormal and uncontrollable growth (McBean, 2012). Like many other post-emergence herbicides, the uptake of dicamba and triclopyr can be improved with the use of an adjuvant (Wang & Liu, 2007).
Adjuvants have multiple functions in relation to pesticide efficacy and include spreaders, stickers, penetrants, and drift-retardants that are either included as formulation components and/or as tank mix additives (Wang & Liu, 2007). Non-ionic surfactants, organosilicones, synthetic and crop oil concentrates are adjuvants commonly used in herbicide mixes (Vanhaecke, 2000). Non-ionic surfactants generally reduce the surface tension of the spray mix, among other effects (Wang & Liu, 2007). The crop oil concentrates typically increase the absorption and penetration of herbicides through the plant cuticle (Wang & Liu, 2007). They have also been shown to delay the crystallization of the herbicides on the leaf surface and reduce the volatile and photodegradative loss of some herbicides (Tse-Seng, Kaben, & Thye-San, 2009). Organosilicones generally lower surface tension and contact angle of water on leaves thereby promoting spreading on the leaf surface (Wang & Liu, 2007).
The problems presented by chemical control of landscape-scale tree invasions highlight some fundamental questions pertaining to the strategy of control and prioritization of infestation typologies (isolated, scattered, and dense), as well as optimum application methods and their interaction with biological efficacy of herbicides (Forster & Kimberley, 2015;Forster, Pathan, Kimberley, Steele, & Gaskin, 2014;Nairn, Forster, & Leeuwen, 2016). The efficacy of foliage-applied herbicide depends on many factors which include the physiochemical properties of the active ingredients, structures and concentration of surfactants, and leaf surface characteristics of the plant species (Stock, Edgerton, Gaskin, & Holloway, 1992). To optimize herbicide use, the factors driving or limiting the foliar uptake and translocation need to be understood. The opportunities to optimize synergism of active ingredients to increase biological efficacy also need investigation.
With reference to these issues and a need to refine the rates of herbicides used in operational control of isolated and dense infestations of Pinus species, a series of experiments were conducted on P. contorta to determine the effect of concentration, dose and adjuvants on needle injury, uptake and translocation of triclopyr (the key herbicide used for control of dense infestations of conifers). The key hypotheses tested were that uptake of triclopyr is proportional to concentration, that adjuvants will enhance triclopyr uptake and that translocation of triclopyr is proportional to uptake and inversely proportional to needle injury.
A similar set of experiments were also conducted to determine the effect of dicamba concentration and adjuvants on uptake of dicamba. This study was followed by an experiment that evaluated the interaction of triclopyr and dicamba when applied as a tank mix, with or without an adjuvant.

| Plant material, herbicides, and adjuvants
Wilding P. contorta seedlings (0.50-0.75 m) were collected from field sites between Taupo and Napier, on the North Island of New Zealand in June 2016. The seedlings were potted on location and brought back to the Scion nursery in Rotorua where they were placed outside and watered. These seedlings were used in all uptake and translocation experiments described below. All plants were equilibrated for 2 weeks before use, and treated, under controlled environment conditions as described for each experiment.
Two herbicide products and a number of adjuvants were used in these studies. In all experiments triclopyr was used as Grazon ® (600 g/L triclopyr butoxyethyl ester, Dow AgroSciences) and dicamba was used as Kamba ® (500 g/L dicamba as dimethylamine salt; Nufarm New Zealand). Details of adjuvants are included in Tables 1-4 in appropriate sections.

| Method used for all uptake and translocation assessments
The quantification of uptake and translocation of triclopyr and dicamba in all experiments described below was made using radiolabelled active ingredients. For all treatments, the radiolabelled active ingredient, either 14 C-triclopyr butoxy ethyl ester (specific activity = 481 MBq/mmol) or 14 C-dicamba (specific activity = 1742 MBq/ mmol), was added to the test formulations at a concentration of 4 MBq/µL. To achieve this, appropriate volumes of radiotracer solution were dispensed into microvials and the carrier solvent (dichloromethane) evaporated under nitrogen. Concentrated solutions of product (either Grazon ® or Kamba ® ) were dispensed on top of the radiotracer, sonicated for 2 min and allowed to equilibrate at room temperature for 1 hr. These were then diluted with appropriate concentrations of formulants to prepare the treatments at the  At 24 hr after treatment (HAT) or 7 DAT, the surface of treated needles in all treatments was washed with 100% acetone (4 ml).
This was diluted with water (1:1) and incorporated into scintillation solution (13 ml ACS, Amersham) to quantify unabsorbed radiolabelled active ingredient. Uptake was determined as the quantity of radiotracer applied minus that recovered in washes. For determination of translocation, treated needles from the 7-day harvest were stored frozen until combusted in a Harvey Oxidiser to quantify radiolabelled herbicide (triclopyr or dicamba) present in the treated needle. The amount of radiolabelled active ingredient not recovered in surface washes or treated needles was determined as translocated. Any loss due to volatilization was not accounted for.
Translocation is reported as a percentage of the amount of herbicide absorbed into the plant.
Where required, the effect of herbicide treatments on needle health was assessed at 24 HAT and/or 7 DAT by scoring injury (browning) on a scale of 0-5, where 0 = no injury and 5 = total brownout of the treated needles.

| Quantification of droplet spread
An assessment of the effect of pine needle surface (inner vs. outer) on droplet spread was made prior to all experiments. Solutions of triclopyr at rates equivalent to 18 kg a.i. ha −1 in 400 L water (or 4.5%) were prepared as described above but with the addition of a water-

| The effect of triclopyr concentration on uptake
The effect of concentration (or dose) on the uptake of triclopyr was determined by the addition of radiolabelled triclopyr to spray formulations of different concentrations (0, 0.75%, 1.5%, 3%, 4.5%, and 6.0%) or use rates (0, 3, 6, 12, 18, and 24 kg a.i. ha −1 ). All treatments were applied at the equivalent of 400 L/ha total spray volume and replicated on the inner and outer surfaces (five droplets on each surface) of ma-

| The relationship between needle injury, uptake, and translocation of triclopyr
A study was carried out to determine the effect of needle injury on uptake and translocation of triclopyr in P. contorta needles when applied at 18 kg a.i. ha −1 as Grazon ® with: (a) no adjuvant, (b) an MSO (5%), and (c) the adjuvants included in the operational "TDPA" mix (

| The effect of concentration and adjuvants on needle injury and uptake of dicamba
Treatments are described in Table 4. Treatments were applied to the mature needles of eight separate P. contorta plants kept under controlled environment conditions as described previously.

| The uptake of triclopyr and dicamba when applied as a tank mix
The final study assessed the interaction between triclopyr and dicamba when applied as a tank mix (as is currently recommended for the operational control of dense infestations of P. contorta; Table 4).
For this study, duplicate solutions containing either 14 C-triclopyr butoxy ethyl ester or 14 C-dicamba were applied as separate treatments on the same replicate trees at the same time, with all other experimental procedures equivalent to that described above. The uptake (7 DAT) of dicamba and triclopyr was determined alone at 5 or 15 kg a.i. ha −1 , respectively, and then in ratios of 1:3, 1:2, and 1:1, with or without an MSO. Needle injury was scored at 1 DAT and at 7 DAT. Treatments were applied to the mature needles of eight separate P. contorta plants kept under controlled environment conditions as described previously.

| Analysis
Data for each experiment were analyzed using the general linear

| Quantification of droplet spread
There was a significant difference (Students t Test; T-value = 9.81, DF = 11; p < .05) in spread of droplets, with significantly more spread on inner needle surfaces (Figure 1). In all further experiments, treatment droplets were placed equally on inner and outer needle surfaces.

| The effect of concentration on uptake of triclopyr
At 1 DAT there were no significant differences in triclopyr uptake, as a percentage of the amount applied, over the concentration range of 0.75%-4.5% (equivalent to dose range of 3 kg to 18 kg a.i. ha −1 ), although the trend was for decreased uptake as concentration increased (Figure 2a). When applied at 6%, or 24 kg a.i. ha −1 , uptake of triclopyr at 1 DAT significantly decreased as a percentage of active ingredient applied (Figure 2a). While a high proportion of triclopyr uptake occurred within 1 DAT, uptake continued up to 7 DAT, especially at the lower concentrations tested (Figure 2b).
Lower uptake at higher concentrations was also observed in similar, preliminary studies with triclopyr covering a concentration range of 3%-6% (Figure 2a). The highest uptake of triclopyr into the needles was achieved with the 18 kg a.i. ha −1 treatment (Figure 2b). Injury to needles was observed to increase with dose but was generally insignificant, with the highest injury score of 1.7 observed for the highest dose (24 kg a.i. ha −1 ) or 6% (data not shown).

| The effect of adjuvants on uptake of triclopyr
In all experiments a high variability between replicate treatments was observed likely caused by the large natural differences occurring in the non-clonal wilding trees used in the study.
Although the highest uptake was obtained with triclopyr alone and the lowest with a MO adjuvant, there were no significant differences between treatments, including the operational treatment (DF = 6; F = 0.13; p = .992). In Experiment 2, the highest uptake of triclopyr occurred where no adjuvant, OS adjuvants, or operational adjuvants were used (Table 1). There was significantly lower uptake of triclopyr where an MO adjuvant was used at either 0.5% or 1.0%. Combining data from Experiments 1 and 2 clearly demonstrated the consistent rate of decline in triclopyr uptake as MO concentration increased (Figure 3).
Based on the results from Experiments 1 and 2, the MO treatments were removed from Experiment 3 and replaced with two concentrations of an MSO adjuvant. In Experiment 3, the highest uptake of triclopyr was achieved with the MSO treatments but uptake was not significantly different to any of the other treatments other than the low rate of OS, where uptake decreased. As with the previous experiments, the adjuvants used operationally did not significantly increase or decrease the uptake of triclopyr relative to the formulated product applied alone.
Normalizing all data to the base treatment of 18 kg a.i. ha −1 , triclopyr applied in Grazon ® with no adjuvant allowed a clearer comparison of the relative effect of adjuvants on uptake across the three experiments ( Figure 4). This comparison indicated that the only adjuvant to increase the uptake of triclopyr applied as the formulated product (Grazon ® ) was the MSO.

| The relationship between needle injury, uptake, and translocation of triclopyr
As in previous experiments large variation in replicate trees was evident.
Triclopyr, applied at 18 kg a.i.ha −1 in the equivalent of 400 L water (or 4.5% F I G U R E 1 Droplet spread on inner and outer needle surfaces of P. contorta needles when applied at 18 kg a.i. ha −1 in the equivalent of 400 L total volume water F I G U R E 2 (a and b) Uptake of triclopyr at 1 DAT and 7 DAT applied as Grazon ® shown as a function of (a) concentration and (b) dose. The points on (a) labeled Prelim 1 and 2 indicate outcomes of similar previous preliminary studies F I G U R E 3 Uptake of triclopyr at 18 kg a.i. ha −1 in 400 L water, or 4.5%, formulated as Grazon ® as a function of inclusion of a mineral oil adjuvant in the mix. Symbols represent data from Experiment 1 (circles) and Experiment 2 (triangles) concentration), was moderately absorbed by foliage (39.8%) within 1 DAT, with 83% of total uptake (7 DAT) occurring within 24 hr (Table 2). By 7 DAT 48% of the amount applied was absorbed. The addition of 20 L (5%) MSO to the carrier increased the rate of uptake within 1 DAT; however, by 7 DAT uptake was not significantly different to that where no adjuvant was used (Table 2). By comparison, uptake at 1 DAT and 7 DAT was significantly lowered where the adjuvants used in the operational TDPA mix were used (DF = 2; F-value = 4.5, p = .024; Table 3). Translocation of absorbed triclopyr was highest (91%) where no adjuvants were used, with a statistically similar high rate of translocation (77%) recorded where an MSO was used ( Table 2). Translocation of absorbed triclopyr was significantly lower (46%) where triclopyr was applied together with the adjuvants used in the operational mix. No injury to needles was observed when triclopyr was applied alone (Table 2). However, adjuvants significantly increased injury, with most of this apparent within 1 DAT. Lower translocation of absorbed triclopyr was associated with higher needle injury, particularly where the operational adjuvants were used.

| THe effect of concentration and adjuvants on needle injury and uptake of dicamba
Uptake of dicamba, applied at 5 kg a.i. ha −1 , was slow and reached 31% at 7 DAT (Table 3). Doubling the dicamba concentration halved the uptake at 7 DAT to 14.3%, effectively delivering the same dose into the plant. Use of MSO at the equivalent of 20 L/ha, or 5%, significantly increased the rate and total amount of uptake of dicamba, as did the adjuvants used in the operational TDPA mix (Table 3). Both of these treatments more than doubled the uptake of dicamba, relative to that where it was applied alone at 5 kg a.i. ha −1 . All treatments where adjuvants were used increased needle injury. Needle injury was low when dicamba was applied alone at either 5 kg a.i. ha −1 or 10 kg a.i. ha −1 .

| The uptake of triclopyr and dicamba when applied as a tank mix
The uptake of dicamba and triclopyr from individual formulations was similar to that observed in previous experiments, 31% and 26%, respectively, at 7 DAT (Table 4). The addition of triclopyr to dicamba treatments significantly increased dicamba uptake at 7 DAT (Table 4) and the ratio of triclopyr to dicamba had no significant effect on this apparent synergism ( Table 4). The addition of dicamba to triclopyr treatments had no significant effect on the uptake of triclopyr ( Table 4). The addition of the MSO at 5% significantly increased the uptake of both active ingredients, particularly where triclopyr was applied at a ratio of 3:1 to the rate of dicamba (dicamba + triclopyr + MSO, 5 + 15 + 5%).
Needle injury for all dual-herbicide treatments was substantially greater than that where herbicides were applied alone, and especially where adjuvants were used (Table 4), as was consistently observed throughout all experiments. The use of the MSO was moderately phytotoxic yet promoted the most herbicide uptake. In earlier studies, such effects were generally not detrimental to triclopyr translocation, but their effects on dicamba movement in the plant, and overall efficacy of treatments were not tested in this study.

| D ISCUSS I ON
The results from these experiments highlighted several factors with respect to the use of triclopyr and dicamba for management of P.
contorta. Firstly, the spread test is a measure of how well a spray can be expected to cover the plant surface following droplet adhesion.
Pinus contorta foliage is classified as moderately easy-to-wet with the inside surface of the fascicle easier to wet than the waxier, outer fascicle surface (Forster et al., 2014), as was illustrated in this study ( Figure 1). In all experiments, treatments were randomly allocated to needle surface and there was no attempt to account for differences in uptake across different surfaces. There is clearly potential to improve treatment spread substantially on the outer needle surfaces, but the effect of that on spray adhesion and uptake into P. contorta, and ultimately efficacy, is unknown.

| Effect of concentration on uptake
One of the factors that effects foliar uptake of active ingredients is the concentration in the spray mix (Wang & Liu, 2007). While concentration is expected to have an important influence on foliar uptake, its effect on uptake of many agrochemicals is largely F I G U R E 4 Uptake of triclopyr applied at 18 kg a.i. ha −1 in 400 L water (18 kg_), formulated as Grazon ® , with inclusion of adjuvants in the mix shown as relative to that where no adjuvant was used. Data across experiments (Table 1)  unknown (Wang & Liu, 2007). Furthermore, its effect on uptake can be dependent on droplet size and number (Huang, Campbell, Studens, & Fleming, 2000), an aspect which was not examined in this study. This study indicated that for triclopyr applied as an ester to P. contorta needles, uptake generally decreased with increasing concentration, particularly above 3%. This resulted in the highest uptake dose being achieved at the operationally applied 18 kg a.i. ha −1 in 400 L total volume water, an outcome that generally supports current operational practice for control of dense infestations of conifers. This outcome contrasts that which has been found for glyphosate, for example, where uptake increased with concentration (Wang & Liu, 2007). In a study that examined absorption and translocation of triclopyr ester in Populus tremuloides, Huang et al. (2000) found that absorption and translocation of triclopyr decreased as concentration increased and droplet size decreased. Absorption and translocation also decreased as droplet number decreased and droplet size increased (while concentration was held constant). They attributed decreases in uptake at higher concentrations to contact injury.

| Effect of adjuvants on uptake
Adjuvants play an important role in increasing the foliar uptake of herbicides (Wang & Liu, 2007). Uptake and translocation of triclopyr applied as Grazon ® in an aqueous carrier in this study were equal to, or in some cases better than, that where adjuvants were added to the formulation. Out of all the adjuvants tested for triclopyr, the MSO applied at between 0.5% and 5% of the carrier volume was the most effective, increasing the rate at which triclopyr was absorbed but not the overall amount, when compared to no adjuvant. The use of adjuvants, notably an MSO at 5% or the adjuvants used in the operational TDPA mix, significantly improved the uptake of dicamba into P. contorta needles, more than doubling the rate of uptake as well as the total amount absorbed 7 DAT. High surfactant oil concentrate adjuvants have been shown to enhance both lipophilic herbicides (triclopyr) and hydrophilic (dicamba) herbicides (Wirth & Zollinger, 2018). In separate studies, While contact injury with high uptake rates of triclopyr was observed in this study, lower uptake into P. contorta at higher concentrations of triclopyr may also be due to a precipitate forming on the needle surface. Lipophilic active ingredients have been shown to form a precipitate on leaf surfaces at concentrations above a certain threshold, dependent on the co-formulants or adjuvants used in the product (Forster & Kimberley, 2015). The crystalline deposit reduces the potential for uptake as penetration from the precipitate can be slow or none at all (Zhu & Lin, 2016). While triclopyr is not a highly lipophilic active ingredient (Log p = .42), the high concentrations being tested in this study could represent the threshold concentrations for triclopyr precipitation.

| Herbicide interactions
The dual-label study indicated that application of triclopyr and dicamba together did not negatively affect uptake of either active ingredient, and in fact triclopyr enhanced the uptake of dicamba.
Increased uptake was also observed when these two active ingredients were applied together with 5% MSO in the equivalent of 400 L water. This result indicates that there is scope to reduce the amount of adjuvant used in the operational TDPA mix without reducing uptake of either active ingredient (and potentially without compromising efficacy).
Ammonium sulfate is currently included at 2.3 kg a.i. ha −1 in the operational TDPA mix. This compound has been shown to increase the efficacy of many weak acid herbicides, including dicamba and picloram, particularly when application is made in hard water, as the sulfate ions bind to the antagonistic cations in the spray solution (Wilson & Nishimoto, 1975;Zollinger et al., 2010). Furthermore, it has been proposed that the ammonium ion increases absorption of weak acid herbicides by increasing their movement across leaf cuticles. However, recent concerns over drift of dicamba in many parts of the United States have highlighted the potential risks of applying dicamba together with ammonium sulfate. When dicamba reacts with spray water above 5.5 pH it forms the anionic form of dicamba which has a low vapor potential; however, below pH 5.5 the dicamba converts to the acid form which has a very high vapor potential (Zollinger et al., 2010). Ammonium sulfate is known to reduce the pH of water (Wilson & Nishimoto, 1975), thereby affecting the volatility of dicamba. Mostly this will not be a problem when spraying wilding pines in remote areas where drift onto neighboring crops is unlikely (and where volatilization may actually contribute to efficacy). However, it would be good practice to reduce the potential for drift, if efficacy can be retained by dropping this compound from the operational TDPA mix. Furthermore, Zollinger et al. (2010) found that use of an MSO increased efficacy of dicamba far more than that of ammonium sulfate. The MSOs increased absorption by dissolving the cuticle and allowing greater penetration of active ingredient than surfactant adjuvants (and petroleum oils; Zollinger et al., 2010).

| CON CLUS IONS
The results indicated that uptake of triclopyr, the key active ingredient used for operational control of wilding conifers, decreased with increasing concentration, with the highest uptake dose achieved at the equivalent of 18 kg a.i. ha −1 triclopyr. The only adjuvant to increase the uptake of triclopyr relative to no adjuvant was an MSO.
The MSO was also shown to double the uptake of dicamba over that where no adjuvant was used. The adjuvants (2.3 kg NH 4 SO 4 , 20 L MSO, 0.5 L surfactant) used in the operational TDPA mix did not increase the uptake and translocation of triclopyr or dicamba over that where only an MSO was included in the mix. This result indicates that there is potential to decrease the amount of adjuvant used without compromising the uptake of triclopyr or dicamba. Needle injury was observed in all studies, and was associated with slightly lower translocation; however, the role of needle injury in final efficacy is unknown. These results indicate that as a first step field studies should test more simple mixes of active ingredients (dicamba, triclopyr, and MSO) and lower rates at similar high volumes (400 L/ ha) in aqueous carriers.