Satsuma Orange Tolerance to Spring and Autumn Indaziflam Applications in Georgia

Abstract

Citrus is a major crop in the SE US, with groves located primarily in Florida, but adapted cultivars have allowed for the expansion of commercial production into the Coastal Plains region of Georgia. Indaziflam, a cellulose biosynthesis inhibiting residual herbicide, controls numerous grass and broadleaf weed species. Research conducted in Georgia from 2020 to 2022 determined the optimal rate and tree response to indaziflam applications. Biannual treatments applied in April and November in established satsuma citrus groves included residual herbicides indaziflam, flumioxazin, diuron, pendimethalin, simazine, and norflurazon. The data indicated no negative impact on tree diameter growth over 30 months after application initiation. Indaziflam provided residual activity in the first year with >80% weed control for bermudagrass and pink purslane and >70% of cutleaf evening primrose, cutleaf geranium, and wild radish. Greater than 69% of weed control was maintained with indaziflam after sequential application for 2 years. All other herbicides provided inadequate residual weed control. Indaziflam PRE applied in citrus groves in Georgia can provide growers with a reliable herbicide option that has been proven to be safe for trees and season-long weed control.

1. Introduction

The southeastern United States has a climate that is conducive for the optimal growth of a multitude of perennial crops, with high temperatures and rainfall averaging 127 cm per year [1]. These include grass species for hay, blueberry, blackberry, tree nut, and peach production. Florida citrus includes oranges, tangerines, and grapefruits that can be sold as fresh or processed goods. These three fruits totaled over 650,000 mt for the 2023 Florida growing season [2]. Interest by growers in southern Georgia has increased for citrus production, as this region has a similar climatic condition (Zone 8a) to the panhandle and north Florida. A recent review highlighted the historical introduction and current economic potential benefits of growing satsuma oranges in southern Georgia [3]. Adapted satsuma cultivars with cold hardiness have been developed for the region [4], and further research is being conducted to establish commercial production. This interest may stem from the issue of Huanglongbing (HLB) disease (citrus greening), as it has spread throughout Florida groves since its introduction in 2005 [5].

HLB disease is spread by the Asian citrus psyllid (Diaphorina citri Kuwayama), which feeds on phloem sap and transmits the bacterium Candidatus Liberibacter asiaticus, causing citrus greening disease [5,6]. No resistant citrus varieties currently exist, but investigators from academia and federal agencies are working to develop resistant cultivars [7]. Citrus crops are typically harvested during the winter months, while a lack of pest control throughout the summer months can prove detrimental to the number and quality of fruit produced. A critical component of field and perennial crop production is weed management. Due to perennial crops not being removed from the field during harvest, tillage practices are not possible, causing a heavy reliance on herbicides for weed control. Investigators have reported that the Asian citrus psyllid can use weed species as a short-term alternate host if citrus plant species are not available for feeding [8].

Indaziflam is registered for multiple perennial fruit and tree nut crops, including citrus [9]. It prevents weed growth and development by disrupting the production of cellulose, which is composed of β-1,4-glucan chains that provide structure to plant cell walls [10]. These cellulose chains are formed on the plasma membrane by the hexagonal cellulose synthase complex protein and must contain at least three cellulose synthase A (CESA) proteins per cellulose synthase complex to form the cellulose microfibrils [11,12]. These microfibrils are layered orthogonally to each other and connect to adjacent microfibrils by cross-linking glycan strands within a pectin network [13]. Indaziflam inhibits cellulose production by increasing CESA density along the plasma membrane paired with reducing CESA particle velocity up to 65%, therefore preventing polymerization [14,15].

Research has been carried out to investigate the efficacy and safety of indaziflam in multiple perennial crops [14,16,17]. Though indaziflam has been indicated to cause minimal injury, certain soil types and agricultural practices may promote injury on trees. Injury was observed in the pecan orchards of Arizona and New Mexico potentially due to tillage practices and soil-to-sand content [18]. Necrotic leaves and various trunk injuries were noted among the affected trees. Southern Georgia’s soils may be composed of more than 90% sand, which may raise concerns for citrus growers.

Though minimal-to-no injury has been reported after indaziflam usage in perennial crops, the response of citrus crops in Georgia has not been evaluated [18,19,20,21]. Therefore, research was performed to establish information about indaziflam for southern Georgia’s citrus growers. The objective was to determine the effects on citrus tree trunk diameter over time and residual weed control activity in citrus production.

2. Materials and Methods

2.1. Site Description

Field studies were conducted in 2-year established (Experiment 1) and newly transplanted (Experiment 2) “Brown Satsuma” citrus trees in Tift County, Georgia (31°34′1.63″ N, 83°36′0.83″ W) from 2020 to 2022. Experiment 1 was initiated in April 2020 and completed in November 2021. Experiment 2 was initiated in April 2021 and completed in November 2022. The trees in Experiments 1 and 2 were within the same field, but different trees were tested in each experiment. The soil samples were collected and analyzed by the Soil, Plant, and Water Laboratory of the University of Georgia (University of Georgia, Athens, GA, USA) and determined to consist of Tift loamy sand (fine-loamy, kaolinitic, thermic Plinthic Kandiudults); pH 6.10; and 83.4%, 9.1%, 7.5%, and 0.75% sand, silt, clay, and organic matter, respectively. The trees were hand planted and maintained using standard citrus agronomic techniques (such as fertilization, irrigation via micro-emitters, and pest control) determined by the grower for the duration of experiments. The trees were deflowered each year to promote vegetative growth.

2.2. Experimental Setup

The experimental design consisted of a randomized complete block with 10 treatments and two application dates, one in April and another in November each year (

Table 1. Application dates for the evaluation of herbicides in citrus trees in Tift County Georgia, 2020 to 2022.

	2020		2021		2022
	1st	2nd	3rd	1st	2nd	3rd
Experiment 1	6 April	4 November	2 April	___	___	___
Experiment 2	___	___	___	2 April	3 November	21 March

), with three replications in Experiment 1 and four in Experiment 2. The average monthly temperatures and monthly total rainfall were recorded (

Table 2. Temperature and rainfall ^a for Tift County in Georgia for 2020, 2021, and 2022.

Year	Month	Maximum Temperature b	Minimum Temperature b	Rainfall c
		_________________________________°C___________________________________		________cm________
2020	April	25.5	16.6	14.2
	May	29.4	19.9	6.6
	June	30.8	20.8	12.9
	July	33.7	22.3	4.7
	August	33.3	22.5	11.6
	September	29.4	20.1	13.3
	October	28.3	17.3	1.4
	November	22.6	12.5	11.8
	December	21.5	11.4	16.5
	Season			93.0
2021	January	15.1	5.7	17.8
	February	17.1	6.4	22.3
	March	23.3	11.1	11.9
	April	23.9	11.6	17.3
	May	28.7	15.7	2.7
	June	31.1	21.4	11.8
	July	31.7	22	20.7
	August	31.9	22.5	14.9
	September	30.4	19.2	9.0
	October	27.8	16.9	5.1
	November	20.1	6.2	1.3
	December	21.2	10.5	7.1
	Season			141.9
2022	January	15.5	3.7	16.7
	February	19.1	6.9	6.0
	March	23.6	10.5	10.1
	April	25.0	12.3	5.8
	May	30.4	18.0	3.2
	June	34.2	21.8	9.9
	July	32.9	22.4	14.2
	August	32.2	21.9	20.7
	September	29.6	18.6	6.3
	October	25.1	10.7	2.6
	November	20.8	10.4	8.7
	December	16.7	6.0	3.6
	Season			107.8

^a Temperature and rainfall data from http://www.weather.uga.edu/ (accessed on10 January 2023). ^b Average of daily values for time period listed over the course of the two experiments. ^c Sum of daily values for each time period listed.

). The treatments consisted of glufosinate at 1269 g ai/ha in combination with either indaziflam at 51 g ai/ha, flumioxazin at 215 g ai/ha, diuron at 1774 g ai/ha, pendimethalin at 2448 g ai/ha, simazine at 2369 g ai/ha, norflurazon at 1123 g ai/ha, or mixed with pendimethalin and simazine at these same rates, with additional treatments of glyphosate alone at 1336 g ae/ha or combined with indaziflam at 51 g ai/ha and a non-treated control (10 total treatments). A glufosinate burndown application (1269 g ai/ha) was made to all plots two to four weeks prior to experimental treatment applications. All treatments included 0.25% v/v of 28% urea ammonium nitrate. Treatments were applied using a CO₂-pressurized backpack sprayer at 187 L/ha with 207 kPa of pressure, utilizing TeeJet TTI11002 nozzles (TeeJet Technologies LLC, Springfield, IL, USA). Applications were made to either side of the plot using a 4 nozzle and 1.8 m boom on the vegetation free strip. The lack of translocation within the plant allowed glufosinate to be used; if it should contact the crop foliage, localized injury would occur and not terminate the crop compared to glyphosate [22]. Glyphosate was used in some treatments for comparison purposes. Plots were 3 m wide by 9 m long containing five trees, with data being collected from the entire plot for residual activity and each tree for trunk measurement.

2.3. Data Collection

Citrus tree trunks were marked with a white paint marker approximately 30 cm above the soil level and above the graft to ensure measurements occurred at the same location on each tree over time. Trunks were measured (cm) using digital calipers angled parallel to the row at the experiment’s initiation and termination on the same mark. The percentage of growth was determined for each tree, then averaged over the five trees, and then averaged over replications using the percent change equation:

$$C=\left({\displaystyle \frac{x_2-x_1}{x_1}}\right)\ast 100$$

(1)

where x₁ indicates the trunk diameter before the treatment application, and x₂ indicates the trunk diameter at the experiment’s termination.

The data collected included visual percent injury (chlorosis or necrosis) compared to the non-treated control (NTC) on a 0 to 100% scale (0% indicating no injury and 100% representing tree death), plant population, and visual percent weed control (0% indicating no control and 100% as total weed control). The citrus groves had sufficient weed pressure at the start of each experiment that consisted of Raphanus raphanistrum L., Oenothera laciniata Hill, Cynodon dactylon L., Portulaca pilosa L., and Geranium dissectum L. to rate the overall weed control.

2.4. Data Analysis

The data were combined across weed species to include the ratings for each respective application timing. All data were subjected to ANOVA to determine the season (spring or autumn) by year interactions. The visual ratings of percent injury, residual control, plant population, and caliper measured trunk diameters were analyzed using PROC GLIMMIX in SAS software (version 9.4, SAS Institute Inc., Cary, NC, USA). Replication was considered a random effect for analysis. Means were separated using Tukey’s HSD at the α < 0.05 level. The trunk diameter data consisted of the combined average of trunks within each plot. The yield data were not collected due to the grower not allowing trees to fruit.

3. Results and Discussion

The initial analysis indicated that the year and experiment were significant, preventing data combination across the year or experiment (data not shown, p < 0.05). Therefore, the data are presented by experiment and year. The data utilized for analysis included visual plant injury, plant population, and weed control compared to the NTC. The data for the tree diameter are presented from the experiment’s initiation to termination.

3.1. Experiment 1

There were no visible signs of injury from any herbicide treatment across the three application timings, nor was there any reduction in stand counts, i.e., no trees died (data not shown). Tree trunk diameters were from 15.8 to 20.5 mm at the experiment’s initiation and 48.0 to 56.4 mm at termination (

Table 3. Citrus tree trunk diameter response to April and November residual herbicide applications.

Treatment	Rate	Experiment 1 a,c						Experiment 2 a,c
		April 2020 b		November 2021		Growth Change		April 2021		November 2022		Growth Change
	_______g ai ha−1_______	___________mm___________				____%_____		__________mm___________				____%____
Nontreated control		18.7	a b	48.0	a	158	a	28.2	a	39.6	b	40.2	a
Glyphosate	1336	19.4	a	50.3	a	160	a	28.8	a	43.0	a	50.6	a
Glyphosate plus indaziflam	1336 + 51	17.6	a	49.3	a	189	a	29.0	a	45.5	a	58.5	a
Glufosinate plus indaziflam	1269 + 51	18.2	a	52.2	a	188	a	30.1	a	44.4	ab	48.3	a
Glufosinate plus flumioxazin	1269 + 215	20.5	a	56.4	a	175	a	30.3	a	43.8	ab	44.5	a
Glufosinate plus diuron	1269 + 1774	15.8	a	48.9	a	213	a	30.0	a	41.6	ab	39.8	a
Glufosinate plus pendimethalin	1269 + 2448	18.1	a	48.3	a	169	a	30.5	a	42.3	ab	38.8	a
Glufosinate plus simazine	1269 + 2369	20.4	a	55.3	a	171	a	30.8	a	43.9	ab	42.5	a
Glufosinate plus pendimethalin plus simazine	1269 + 2448 + 2369	18.5	a	51.9	a	181	a	29.2	a	42.4	ab	45.9	a
Glufosinate plus norflurazon	1269 + 1123	19.6	a	50.0	a	158	a	29.8	a	43.0	ab	44.2	a

^a ANOVA using the GLIMMIX procedure in SAS, Type 3 Tests of Fixed Effects (SAS Institute, Cary, NC, USA) prevented data combination. ^b Letters indicate statistical significance at the α < 0.05 within each season and year, according to Tukey–Kramer HSD. ^c Experiment 1 consisted of 2-year-old trees and Experiment 2 consisted of newly transplanted trees within the same field.

). No differences for the treatments were noted at the experiment’s initiation or termination. Trees did not sustain a significant reduction in growth due to herbicide application, as all treatments were equal to or greater in size compared to the NTC. The change in growth was also determined and indicated to be greater than a 158% increase over time for all treatments. The least amount of growth occurred in the NTC, and for glufosinate plus norflurazon, at 158%. Overall, the greatest tree growth occurred where glufosinate plus diuron was applied, but no differences for any treatment were indicated. Trees in plots treated with glyphosate or glufosinate plus indaziflam indicated the next greatest amount of growth at 189% and 188%, respectively. Previous research with pecan over a three-year period exhibited similar results, and there was no visible injury or stand loss from repeated indaziflam application [16].

Ratings for percent residual control were recorded after the first glufosinate application made after treatments until the next burndown application (

Table 4. Residual weed control one month after herbicide application in Tift County Georgia for two-year-old citrus trees, Experiment 1 ^a.

Treatment	Rate	April 2020		November 2020		April 2021
	___g ai ha−1___	___________________________%___________________________
Nontreated control		0 b	c c	0	e	0	e
Glyphosate	1336	39	b	26	d	22	de
Glyphosate plus indaziflam	1336 + 51	80	a	88	a	89	a
Glufosinate plus indaziflam	1269 + 51	80	a	69	ab	88	a
Glufosinate plus flumioxazin	1269 + 215	92	a	64	b	73	ab
Glufosinate plus diuron	1269 + 1774	84	a	81	a	56	a–c
Glufosinate plus pendimethalin	1269 + 2448	72	ab	33	cd	47	b–d
Glufosinate plus simazine	1269 + 2369	76	ab	41	c	28	c–e
Glufosinate plus pendimethalin plus simazine	1269 + 2448 + 2369	91	a	75	ab	63	ab
Glufosinate plus norflurazon	1269 + 1123	41	b	34	cd	53	b–d

^a Weed species included Raphanus raphanistrum L., Oenothera laciniata Hill., Cynodon dactylon L. Pers., Portulaca pilosa L., and Geranium dissectum L. ^b ANOVA using the GLIMMIX procedure in SAS, Type 3 Fixed Effects (SAS Institute, Cary, NC, USA) prevented data combination. ^c Letters indicated statistical significance at the α < 0.05 within each season and year, according to Tukey–Kramer HSD.

). Greater than 72% control was achieved by all treatments except for the NTC, glyphosate alone, and glufosinate plus norflurazon. Glyphosate alone provided no long-term residual control [22] and did not differ from glufosinate plus pendimethalin, simazine, or norflurazon.

Fall residual control of all treatments decreased 3 to 39% compared to spring treatment residual control (Table 3). Glufosinate plus pendimethalin had the greatest reduction in both residual control and percent control compared to the NTC in the autumn of 2020. This is due to the limited residual activity of pendimethalin, as it dissipated with rainfall over time [23].

Spring 2021 residual control decreased in all treatments except glyphosate or glufosinate plus indaziflam and glufosinate plus norflurazon compared to spring 2020 applications (Table 3). The decrease in residual control ranged between 17 and 48% for the respective treatments, while glyphosate or glufosinate plus indaziflam increased by 9 or 8%, respectively. Glufosinate plus norflurazon residual control increased by 12% compared to the previous spring application. All treatments provided 73% or less residual control, except for the indaziflam tank mixtures that were >88% in May 2021.

3.2. Experiment 2

Experiment 2 was located within the same field as Experiment 1 and contained the same variety of satsuma, but they were newly transplanted and not established. Weed pressure was present at the beginning of Experiment 2 and noted to be denser than Experiment 1. ANOVA indicated differences by season and year, preventing the data from being combined for presentation (

Table 5. Residual weed ^a control one month after herbicide applications in Tift County Georgia in newly planted citrus trees, Experiment 2.

Treatment	Rate	April 2021		November 2021		April 2022
	___g ai ha−1__	_________________________%_________________________
Nontreated control		0 b	d c	0	e	0	f
Glyphosate	1336	30	b-d	16	de	17	ef
Glyphosate plus indaziflam	1336 + 51	62	ab	81	a	79	a
Glufosinate plus indaziflam	1269 + 51	55	ab	75	ab	69	ab
Glufosinate plus flumioxazin	1269 + 215	66	a	39	cd	39	b–e
Glufosinate plus diuron	1269 + 1774	40	a–c	55	a–c	50	a–d
Glufosinate plus pendimethalin	1269 + 2448	28	b–d	34	c–e	37	d–f
Glufosinate plus simazine	1269 + 2369	43	a–c	30	c–e	23	b–e
Glufosinate plus pendimethalin plus simazine	1269 + 2448 + 2369	18	cd	47	b–d	45	b–e
Glufosinate plus norflurazon	1269 + 1123	28	b–d	57	a–c	64	a–c

^a Weed species included Raphanus raphanistrum L., Oenothera laciniata Hill., Cynodon dactylon L. Pers., Portulaca pilosa L., and Geranium dissectum L. ^b ANOVA using the GLIMMIX procedure in SAS, Type 3 Fixed Effects (SAS Institute, Cary, NC, USA) prevented data combination. ^c Letters indicated statistical significance at the α < 0.05 within each season and year, according to Tukey–Kramer HSD.

).

The trunk diameters of trees in Experiment 2 were slightly larger at initiation compared to Experiment 1, ranging between 28.2 and 30.5 mm with no differences noted (Table 3). The diameters at the experiment’s termination ranged between 39.6 mm and 45.5 mm. The differences were indicated in plots treated with glyphosate plus indaziflam, with trees having a greater diameter compared to the NTC. Glyphosate plus indaziflam and glufosinate plus indaziflam were not different from each other, but the misapplication of glyphosate may have more severe consequences than glufosinate due to its systemic properties compared to the contact properties of glufosinate [24,25]. However, when applied correctly, both glyphosate and glufosinate do not adversely affect citrus growth [26,27]. The percentage growth was also lower compared to Experiment 1, ranging from 38.8% to 58.6%. The trees in plots treated with glyphosate and indaziflam noted the largest amount of growth. The reduced trunk growth may be due to the increased weed pressure, lowering the amount of available water and nutrients to the trees.

No treatment provided greater than 66% residual control one month after the April 2021 applications (Table 5). This may have been due to the residual herbicide not contacting the soil surface and remaining on the plant matter due to heavy weed presence or not receiving proper rainfall for activation [28]. Glufosinate plus flumioxazin provided the greatest amount of residual control at 66%, while the tank mixture of glufosinate plus pendimethalin plus simazine provided only 18%. Flumioxazin has been noted to provide adequate weed control in numerous crops [29,30,31].

Autumn herbicide treatments applied in November 2021 gave a residual control of 55% or lower for all treatments, except for glufosinate or glyphosate plus indaziflam. These combinations provided 81% and 75% control, respectively. Unusually warm winter weather experienced following the autumn applications in 2021 may have increased the weed pressure, causing the reduced control.

Spring 2022 residual control one month after the April application noted a decrease in all treatments except glyphosate alone, glufosinate plus pendimethalin, and glufosinate plus flumioxazin, which either remained the same or slightly increased (Table 5) compared to the Autumn 2021 application. Both indaziflam treatments had the greatest residual control, providing 69% or greater, followed by norflurazon at 64% and diuron at 50% residual control.

Reports have indicated that injury may occur from indaziflam applications under specific environmental conditions. Soils that are predominately sand and maintained under flood irrigation have been noted to be conducive to crop injury from indaziflam [18,19]. South Georgia soils typically have sand content above 90% with pH levels between 5 and 6, as noted in these experiments.

Crop root structure may also play a role in the occurrence of injury. Pecan lateral roots are typically within the top 15 to 30 cm; whereas, citrus roots may be as deep as 91 cm, therefore increasing the depth to which indaziflam must travel to cause injury. The label states that citrus trees must be established for greater than one year or be transplanted from pots for longer than one month [9]. The irrigation practices are predominately micro-emitters or non-irrigated, reducing the movement of indaziflam through the soil profile by only allowing small amounts of water be emitted for long periods of time, reducing movement into the soil. This is supported by research [32] in which the investigators reported no indaziflam injury on grape or muscadine growth, yield, or quality when grown under similar soil and irrigation conditions to these citrus experiments [33].

4. Conclusions

The applications of indaziflam in Georgia citrus groves provided increased residual weed control (up to 88%) over other commonly applied herbicides, and when no residual herbicide was applied. The weed control provided by indaziflam application removes host sites for insect pests and allows crop inputs to be utilized by the trees and not weeds. One concern with herbicides applications is crop response in the form of necrosis, chlorosis, or crop stunting. No crop injury was noted in either experiment, with respect to applications of indaziflam. Therefore, herbicides containing indaziflam may be integrated into south Georgia’s citrus production due to its effectiveness and safety.