Integration of Organic Amendments and Weed Management to Improve Young Citrus Tree Growth Under HLB-Endemic Conditions

Abstract

Florida citrus production has declined by over 90% since the bacterial disease huanglongbing (HLB) was found in the state. In the absence of an effective cure, growers are adopting more frequent fertilization and irrigation practices to improve tree health and prolong the life span of their orchards. However, Florida’s soils under citrus production are sandy, with little organic matter, a low water holding capacity, and a low cation exchange capacity (CEC), rendering them prone to nutrient leaching. Organic amendments can be used to improve soil health and the environment for citrus roots, but may promote a higher incidence of weeds competing with trees for water and nutrients. A large field trial was established in a commercial citrus orchard in southwest Florida to evaluate the effects of organic amendments and weed management on young tree growth. The organic amendment treatments were as follows: (1) plant-based compost, (2) humic acid, and (3) a non-amended control. The weed management (herbicide) treatments were (1) glyphosate, (2) glufosinate, (3) flumioxazin, and (4) a maintenance herbicide control. Trees were planted in August 2019, and treatments began in 2021. Tree growth and physiological variables and soil physicochemical properties were evaluated during the two-year study. Compost-amended plots had a higher volumetric water content throughout the experiment, and soil nutrient content, organic matter, CEC, and pH were higher after two years of application. Humic acid amendments were less effective in altering these soil properties. Compost’s effects on tree and fibrous root physiology were moderate, and tree growth, fruit yield and fruit quality were not affected by either organic amendment. In contrast, the use of post-emergent herbicides (glyphosate and glufosinate) improved tree growth and nutrient uptake. The results suggest that in Florida, the use of organic amendments needs to be integrated with weed management to prevent resource competition. In the short term, these practices did not improve the productivity of the trees in the current Florida production environment.

1. Introduction

Citrus production in Florida has declined dramatically in the past two decades, placing Florida second behind California among US citrus-producing states [1,2,3]. The total citrus acreage decreased from 302.9 thousand ha in 2004 to 111.2 thousand ha in 2023–2024 [3], and production declined from 13 million tons to 0.8 million tons [4]. While hurricanes have contributed, the main reason for this drastic decline is huanglongbing (HLB, a.k.a., citrus greening), a bacterial disease first discovered in Florida in 2005 [5] and that has been endemic since 2013 [1]. Despite being severely impacted by HLB, the Florida citrus industry remains a major force contributing to the state’s overall economy [6].

In Florida and most other affected citrus-growing areas around the world, HLB is associated with the phloem-limited Gram-negative bacterium Candidatus Liberibacter asiaticus (CLas), which is transmitted by the Asian citrus psyllid (ACP) Diaphorina citri [7,8]. Infected trees have a wide range of symptoms, including chlorotic and/or blotchy mottled leaves, canopy dieback, small misshapen and poorly colored fruits, poor internal fruit quality, and premature fruit drop [7,9,10].

One of the early consequences of HLB is the loss of fibrous roots, which can commence before foliar disease symptoms are visible in the canopy [11]. The HLB-associated fibrous root loss negatively impacts water and nutrient uptake, hampering overall tree performance [12]. In addition to modifying nutrition and irrigation practices, promoting a healthy root system may help cope with this devastating disease. One way to improve root health is by improving the soil physicochemical properties. A more favorable soil environment can improve root physiological properties and improve plant growth in multiple ways, such as increasing the rhizosphere microbial activities, improving nutrient cycling and availability, and enhancing resilience to biotic and abiotic stresses [13,14].

Florida’s citrus growers have modified their management practices to mitigate the impact of HLB and maintain the productivity of affected citrus orchards under high disease pressure. Major strategies include insecticide applications [15], enhanced nutrition [16], plant growth regulator applications [17], and the use of tolerant rootstocks [18,19]. More recently adopted strategies like the use of individual protective covers (IPCs) to exclude the HLB vector [20] and trunk injection of oxytetracycline to reduce bacterial titers [21] have demonstrated more dramatic improvements in tree health and productivity, but are labor-intensive, costly, and considered short-term management options. Hence, longer-term, more sustainable cultural practices are needed to maintain productivity in the HLB era.

Florida soils under citrus cultivation are typically sandy with very low organic matter (<1%) [22], rendering them ineffective in retaining water and nutrients and promoting leaching. The loss of fibrous roots associated with HLB, in combination with the unfavorable soil properties, poses significant challenges for the Florida citrus industry. Heavy rainfall events, especially during the summer, further exacerbate nutrient leaching, increasing production costs and raising environmental concerns. Results from previous research on citrus and other crops suggest that amending soils with organic matter can help improve overall soil physicochemical properties, including porosity and aggregate stability, pH, moisture, and CEC [23,24,25]. Organic amendments can also enrich the root–soil microbial communities and attract plant growth-promoting rhizobacteria (PGPR), which could assist in improving nutrient availability and nutrient use efficiency and increasing stress tolerance [26,27]. In a recent study, the diversity of citrus rhizosphere microbes related to nutrient cycling was enriched with the application of different organic soil amendments, including compost and fulvic acid, but the study was conducted on mature trees and did not have any impact on tree health and productivity [28].

Compost is a stabilized product of biodegraded organic materials such as manures and plant and animal residues, which can improve nutrient retention and help balance soil nutrients, resulting in a more gradual release [24]. However, because of the variable range of mineralization, compost application may take multiple years to have visible effects on tree growth. Humic acids are organic macromolecules formed by the microbial decomposition of animal and plant residues or chemical extraction [29]. Compared with compost, humic acids are more stable, and their effects are generally considered to be more immediate. The plant response to humic substances varies by species, the mode and rate of application, and existing management and environmental conditions [29].

One drawback of using organic amendments is that the favorable soil environment created by their use may encourage weed growth [30]. Weeds compete with plants for water, nutrients, light, and space, and can act as alternative hosts for different pests and pathogens [31]. Florida’s favorable weather conditions, characterized by high precipitation and warm, humid temperatures, create an ideal environment for weed proliferation year-round, presenting significant challenges in citrus-producing areas. If left unmanaged, weed competition may be detrimental to the establishment of young trees after transplant, slowing growth and ultimately reducing productivity. Herbicides are the most efficient way to control weeds [32]. Glyphosate (N-[phosphonomethyl] glycine), a post-emergence herbicide, ranks number one among the herbicides used for weed management in Florida citrus production [33]. It controls annual grasses and broad-leaf weeds by disrupting the shikimic acid pathway (HRAC/WSSA mode of action group 9). Glufosinate (ammonium 2-amino-4(hydroxymethylphosphinyl) butanoate), another systemic post-emergent herbicide, is emerging as an alternative to glyphosate [34]. Glufosinate inhibits glutamine synthetase activity (HRAC/WSSA mode of action group 10) to kill weeds. Flumioxazin (2-[7-fluoro-3,4-dihydro-3-oxo-4-(2-propynyl)-2H-1,4-benzoxazin-6-y]-4,5,6,7-tetrahydro-1Hisoindole-1,3(2H)-dione) is a pre-emergence herbicide with limited information available regarding its application in citrus weed management. Flumioxazin can efficiently control broad-leaf weeds and grasses in other crops by inhibiting chlorophyll biosynthesis (HRAC/WSSA mode of action group 14).

Soil-applied fertilizers, irrigation, and weed management account for 26%, 13%, and 14% of the total costs of citrus production in Florida [35]. Organic amendments can increase the economic efficiency of nutrition and irrigation programs by reducing the need for frequent fertilization and irrigation. The integration of organic amendments and weed management may help to improve soil health and plant performance while reducing the cost of production. Despite the widespread use of organic amendments in crop production, there is limited information on the interaction with weed management, especially in citrus. Therefore, this study aimed to evaluate the response of young citrus trees to repeated amendments of compost and humic acids and the interaction with different weed management practices in a commercial production system. We hypothesized that repeated applications of organic soil amendments would improve soil physicochemical properties, supporting tree establishment and promoting root health and tree growth, but that weeds need to be managed adequately to maximize efficiency under Florida growing conditions.

2. Materials and Methods

2.1. Study Site and Plant Material

A large-scale field study was conducted in a commercial citrus orchard near Felda, Hendry County, FL, USA (26°34′27.9″ N 81°30′38.5″ W). The soil at the study site was a poorly drained sandy spodosol [22]. Soil analysis at the beginning of the study showed that it was composed of 92.5% sand, 2.5% clay, and 5% silt, with an organic matter content of 0.5%, a pH of 6.7, and a cation exchange capacity (CEC) of 3.9 meq/100 g. The soil nutrient content was 42.1 ppm phosphorus, 11.5 ppm potassium, 41.5 ppm magnesium, 512 ppm calcium, 0.5 ppm nitrate, and 0.9 ppm ammonium. The trees used for the study were ‘Valencia’ sweet orange (Citrus sinensis) grafted on US-802 (C. maxima × Poncirus trifoliata) rootstock. Trees were planted in 2019 in double rows on raised beds separated by furrows at a spacing of 3.7 m within rows and 7.6 m between rows. Treatment application commenced in 2021. Since HLB is endemic in Florida [1], all trees were affected by HLB at the beginning of the study and exhibited moderate foliar disease symptoms. The average daily temperature during the experiment was 23.5 °C, and total rainfall was 3355 mm, varying from 4.3 mm in December 2021 to 445.3 mm in June 2022 (Figure 1).

2.2. Experimental Design and Treatments

The experimental design was a completely randomized 3 × 4 split-plot design with organic amendment as the main plot factor and weed management (herbicide) as the subplot factor. There were 3 organic amendment treatments—(1) compost, (2) humic acid, and (3) no amendment (control)—and four herbicide treatments: (1) glyphosate, (2) glufosinate, (3) flumioxazin, and (4) maintenance herbicides (control; see below). Plots were arranged in 5 blocks, each containing 3 beds, 1 bed for each type of organic amendment. Each bed contained 200 trees arranged in two rows of 100 trees with subplots of 50 consecutive trees for each herbicide treatment. In total, there were 60 experimental plots (5 blocks × 3 organic amendment treatments × 4 herbicide treatments). An aerial view of the experimental site is shown in Figure 2.

Treatments commenced in May 2021. Compost was applied at 12.3 tons/ha per application by broadcast spreading underneath the tree canopy. Humic acid was applied at a rate of 0.45 kg/tree by hand underneath the tree canopy. The compost used in the experiment was derived from yard waste and was locally sourced (Kastco Agriculture Service, Naples, FL, USA). It had the following physicochemical properties: 51.1% total solids, 23.6% organic matter, pH 7.7, conductivity 3.1 mS/cm, 0.52% total nitrogen (0.52% organic nitrogen, 0.001% ammonium nitrogen, <0.01 nitrate nitrogen), 0.08% phosphorus, 0.18% phosphorous as P₂O₅, 0.26% potassium, 0.31% potassium as K₂O, 0.09% sulfur, 3.28% calcium, 0.31% magnesium, 0.07% sodium, 2500 ppm iron, 67.5 ppm manganese, and <100 ppm boron. The humic acid was a dry, granular material (Soil Boost, Soil Biotics, Kanakakee, IL, USA) containing 70% humic acids.

Glyphosate (tradename: Roundup^®, Bayer Crop Science, NC, USA) was applied at 4.3 kg a.i/ha. Glufosinate (brand name: Interline^®, UPL NA Inc., King of Prussia, PA, USA) was applied at 1.4 kg a.i/ha. Flumioxazin (brand name: Chateau^®, Valent LLC, San Ramon, CA, USA) was applied at 0.3 kg a.i/ha. Herbicide rates were based on the recommended rates for citrus in this region and were within the range specified on the label. Herbicides were dissolved in water, and a water conditioning agent (Quest^®, Helena Agri-Enterprises, Collierville, TN, USA) and surfactant (Induce^®, Helena Agri-Enterprises, Collierville, TN, USA) were added according to the manufacturer’s recommendations. Herbicides were applied underneath the tree canopy using a boom sprayer mounted on the front of the tractor and a spray volume of 310 L/ha. Organic amendments and herbicides were applied twice a year, in May and November. As maintenance herbicides to manage weed growth in the untreated control plots, a standard grower mixture of saflufenacil (brand name: Treevix^®, BASF Corporation, Florham Park, NJ, USA) and sethoxydim (brand name: Poast^®, BASF Corporation, Florham Park, NJ, USA) was applied at the recommended rates twice a year. Irrigation was performed by seepage from the water furrows in between the citrus beds. Other management practices were conducted per the growers’ standard practices. During the study period, a freeze event occurred on 29 and 30 January 2022, and a hurricane (hurricane Ian) crossed Florida on 28 September 2022, causing some defoliation and fruit drop. The study had to be terminated in spring 2023, after the harvest, as the grove operator ceased all citrus production in this area.

2.3. Tree Growth

Unless stated otherwise, analyses were conducted on 10 random trees (“flagged trees”) per plot, five in each of the opposite rows.

Tree size differed significantly among plots at the beginning of the study (Table S1). Therefore, tree growth was determined as the percent increase in tree height, canopy volume, and scion trunk diameter from November 2021 to November 2022. Tree growth was assessed by measuring tree height, scion trunk circumference, and canopy width. Scion trunk circumferences were measured 5 cm above the graft union, and trunk diameters were determined using the circle circumference (C) formula C = π × diameter. Tree height was measured from the soil surface to the top of the trees, excluding erratic shoots. The canopy width was measured in two perpendicular directions and averaged. The canopy volume was calculated using the formula by Wutscher and Hill [36]: canopy volume = (width² × height)/4.

2.4. Candidatus Liberibacter asiaticus (CLas) Detection

Four mature leaves from the most recent flush were randomly collected from three of the flagged trees in each plot in November 2022, and pooled. Fibrous roots (<2 mm in diameter) were collected from the upper 20 cm of the soil beneath the tree canopy from the same trees on the same day and pooled. Leaf and root tissues were stored at −20 °C until analysis. Root tissues were washed with water and blotted dry before storage. Tissues were pulverized with liquid nitrogen using a mortar and pestle, and 100 mg of each sample was used to extract DNA using the Plant DNeasy^® Pro-Kit (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions. Real-time quantitative polymerase chain reaction (qPCR) assays were conducted to detect the presence of CLas 16S rDNA using primers HLBas/HLBr and probe HLBp, and primers COXf and COXr and probe COXp as an internal control and for normalization [37]. Amplification was performed over 40 cycles using the iTaq™ Universal Probes Supermix (Bio-Rad, Hercules, CA, USA) in a QuantStudio 3 real-time PCR system (Applied Biosystems, Foster City, CA, USA) according to the manufacturer’s instructions. All reactions were carried out in a 20 µL reaction volume using 2 µL DNA. Samples were considered CLas-positive if their normalized cycle threshold Ct-values were ≤36.

2.5. Leaf Nutrient Analysis

A random sample of five leaves per tree was collected in July 2022 and 2023 from each of the 10 trees in each experimental plot, and pooled. Samples were sent to Waters Agricultural Laboratories, Inc., Camilla, GA, USA, for macro- (N, P, K, Ca, Mg, S) and micronutrient (B, Zn, Mn, Fe, Cu) analysis using inductively coupled plasma (ICP) emission spectroscopy [38].

2.6. Fruit Drop

Fruits on the ground and retained on the tree were counted on the 10 flagged trees per experimental plot in February 2022. Fruit drop was calculated as the percentage of dropped fruits relative to the sum of dropped and retained fruits. For the 2022–2023 season, fruits under the canopy were counted on 4 trees per experimental plot in monthly intervals from November 2022 to February 2023 (before harvest). After each count, dropped fruits under the canopy were removed with a rake. The total number of retained fruits was counted in February 2023 prior to the harvest, and the cumulative fruit drop percentage was calculated.

2.7. Yield and Juice Quality Analysis

Fruits were counted on the 10 flagged trees per experimental plot in February 2022 and 2023. The average fruit count per tree was multiplied by the average weight per fruit to obtain the yield per tree. For fruit/juice quality analysis, a half-bushel bag of fruits (45–55 fruits) was collected randomly per experimental plot in February 2022 and 2023 just before harvest. The fruits were analyzed at the Juice Processing Pilot Plant, CREC, UF/IFAS, Lake Alfred, FL, USA. Fruit weight, juice percentage, juice color, total soluble solid (TSS—°Brix), percent titratable acid (TA), and the TSS/TA ratio were determined using standard commercial procedures.

2.8. Root Health Assessments

Root health assessments included root density, fibrous root respiration, and specific root length (SRL). To determine the root density, root core samples were collected in August 2022 using a 750 cm³-volume root auger (Eijkelkamp, Wilmington, NC, USA). Four cores were taken from underneath the canopy of two trees in each experimental plot and pooled. Roots were separated from the soil, washed with water, and blotted dry before drying them at 65 °C until constant weight to determine the dry biomass. Root density was calculated by dividing the dried biomass of the roots by the total volume of the core.

The fibrous root respiration rate was measured at the trial site every three months from April 2022 to April 2023. Fibrous roots (<1.5 mm in diameter) were collected using a soil knife from multiple locations (≥3) beneath the tree canopy at 20 cm depth. Roots were washed with water and blotted dry. A sub-sample of fibrous roots was placed in a 50 cm³ respiration chamber (LI-COR Bioscience, Lincoln, NE, USA), and gas exchange was measured for 3 min using a LI-COR LI-850 CO₂/H₂O analyzer. The respiration rate was expressed as mg CO₂ per g of root dry weight per hour (CO₂/g/h). The remaining fibrous roots were kept on ice until transport to the laboratory to determine the SRL. Roots were separated by root order and were scanned at 300 dpi using a flatbed Epson V850 Pro scanner (Epson, Los Alamitos, CA, USA). The total root length was determined using RhizoVision Analyzer Interactive software (version 2.0.2) [39] following the algorithms in Seethepalli et al. [40]. The SRL was calculated for each root order by dividing root length by the dry weight of the scanned roots.

2.9. Soil Properties

The volumetric soil moisture content (%) was measured underneath the edge of the tree canopies of 2 of the 10 flagged trees of each experimental plot using a FieldScout TDR 350 soil moisture meter (Spectrum Technologies, Aurora, IL, USA). Measurements were conducted monthly from March 2022 to April 2023.

An 8 × 21″ (20 × 53 cm) soil probe (AMS, American Falls, ID, USA) was used to collect bulk soil samples from beneath the canopy drip line of the 10 flagged trees at 15 cm depth. Soil samples were sent to Waters Agricultural Laboratories, Inc., Camilla, GA, USA, for analysis of pH, CEC, organic matter, nitrate, and macro0 (N, P, K, Ca, Mg, S) and micronutrient content (B, Zn, Mn, Fe, Cu). Extractions were performed using Mehlich III extraction procedures [41].

2.10. Weed Biomass and Diversity

To evaluate the herbicide efficacy, the aboveground weed biomass was measured in two random quadrats of 0.25 m² near the edge of the canopy in each experimental plot two months before herbicide applications in March 2022, August 2022, and April 2023. Weeds were cut at the soil surface and oven-dried at 65 °C until constant weight. The percentages of grass weeds and broadleaf weeds were determined from weed samples collected in August 2022 and April 2023.

The weed diversity was assessed by counting and identifying the major weeds in two random quadrats of 0.25 m² near the edge of the canopy of two trees in each experimental plot. Weeds were counted one month after each herbicide treatment application in June 2022 and January 2023. Weed diversity has been presented as the percentage of major weed species in the total weed count.

2.11. Statistical Analysis

All statistical analyses were conducted in RStudio Version 4.3.1 [42]. A split-plot two-way analysis of variance (ANOVA) was performed for all variables except soil moisture content. A linear mixed model was fitted for organic amendments and herbicide treatment as fixed factors and block as a random factor using the lmerTest [43]. Soil moisture data were analyzed using a two-way repeated measures ANOVA with month and organic amendment as fixed factors and block as a random factor. Before performing ANOVA, the data were checked for normality and homogeneity of variance. If needed, data were transformed to meet the assumption of ANOVA. The post-hoc comparison of means was conducted using Tukey’s honest significant difference test. Differences were defined as statistically significant when the p-value was smaller than 0.05.

3. Results

3.1. Tree Growth

The average tree height, canopy volume, and scion trunk diameter at the end of the study were 1.73 m, 1.91 m³, and 6.7 cm, respectively, and were not significantly affected by any of the treatments (Table S1). There were no significant differences in the tree growth rate in response to the different organic amendments (

Table 1. Increase in tree height, canopy volume, and scion trunk diameter of ’Valencia’ trees grown in southwest Florida with different organic amendments and herbicides.

	Tree Height Increase (%)	Canopy Volume Increase (%)	Scion Trunk Diameter Increase (%)
Organic amendment
Compost	17.4	81.2	30.3
Humic acid	18.6	79.6	31.6
Control	17.2	73.6	30.8
p-value	0.7447	0.7275	0.4865
Herbicide
Glyphosate	19.0 ab	90.4 a	36.0 a
Glufosinate	20.6 a	81.4 ab	31.68 b
Flumioxazin	15.8 bc	66.6 b	27.6 b
Control	15.6 c	74.3 ab	28.6 b
p-value	0.0001	0.0140	<0.0000
Organic amendment × Herbicide
p-value	0.8899	0.2270	0.2991

Increases were determined as the percent change from November 2021 to November 2022. Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). In contrast, herbicide treatment significantly influenced growth. Trees from glufosinate-treated plots grew significantly more in height (20.6%) than trees from flumioxazin-treated (15.8%) and control plots (15.6%). Trees from glyphosate-treated plots increased their canopy volume more (90.4%) than trees from flumioxazin-treated plots (66.6%). The scion trunk diameter increased more in glyphosate-treated plots (36.0%) compared to glufosinate (31.7%), flumioxazin (27.6%), and control (28.6%) plots. There was no significant interaction between organic amendment and herbicide for any of the growth variables.

3.2. Candidatus Liberibacter asiaticus (CLas) Detection

CLas titers are expressed as cycle threshold (Ct) values. A high Ct value indicates low CLas titers, and a low Ct value indicates high CLas titers. Ct values were 21.4 and 33.4 on average in leaves and roots, respectively, but were influenced by neither the organic amendments nor the herbicide treatment (Table S2).

3.3. Leaf Nutrients

In July 2022, one year after treatment application, organic amendments did not influence the leaf nutrient content, except for potassium (

Table 2. Leaf nutrient contents of ’Valencia’ trees grown in southwest Florida with different organic amendments and herbicides.

	N (%)	P (%)	K (%)	Ca (%)	Mg (%)	S (%)	B (ppm)	Zn (ppm)	Mn (ppm)	Fe (ppm)	Cu (ppm)
July 2022
Organic amendment
Compost	2.37	0.20	1.98 a	2.88	0.31	0.24	61.1	34.8	18.5	73.3	2.9
Humic acid	2.34	0.20	1.55 b	2.84	0.31	0.24	53.7	31.6	18.8	75.3	2.8
Control	2.32	0.20	1.51 b	2.89	0.29	0.23	56.8	29.5	17.4	68.0	2.9
p-value	0.7839	0.8641	<0.0000	0.7696	0.2874	0.7770	0.0587	0.0810	0.3028	0.3579	0.9249
Herbicide
Glyphosate	2.47 a	0.21 a	1.69	2.9	0.31	0.25 a	54.7	34.0 a	19.5 a	71.3	2.9
Glufosinate	2.40 a	0.21 ab	1.73	2.93	0.30	0.24 ab	59.7	32.1 ab	18.3 ab	65.9	2.9
Flumioxazin	2.26 b	0.20 bc	1.68	2.88	0.32	0.24 ab	57.0	31.4 ab	18.2 ab	71.3	2.9
Control	2.23 b	0.19 c	1.63	2.75	0.30	0.22 b	57.1	30.3 b	16.7 b	75.0	2.8
p-value	<0.0000	0.0002	0.1195	0.1257	0.1841	0.0005	0.1491	0.0127	0.0029	0.2791	0.8778
Organic amendment × Herbicide
p-value	0.3377	0.5264	0.9248	0.7688	0.9669	0.4435	0.5698	0.5509	0.6669	0.8747	0.4675
July 2023
Organic amendment
Compost	2.88	0.21	2.37 a	2.81	0.31	0.33	61.8 a	33.2	29.2	61.9	161.3
Humic acid	3.05	0.21	2.07 b	2.95	0.32	0.33	55.31 b	32.9	30.3	59.8	154.5
Control	2.90	0.21	2.16 b	2.87	0.31	0.31	55.0 b	32.1	29.3	59.3	160.6
p-value	0.2141	0.7638	0.0031	0.4210	0.0752	0.5405	0.0044	0.7442	0.7632	0.3644	0.9566
Herbicide
Glyphosate	3.01 a	0.21	2.13	2.91	0.32	0.32	53.8 b	33.9 a	31.1 a	60.3	148.9
Glufosinate	3.02 a	0.21	2.18	2.88	0.31	0.32	57.7 ab	33.7 ab	30.4 ab	61.5	161.3
Flumioxazin	2.93 ab	0.21	2.25	2.87	0.31	0.32	55.5 ab	32.0 ab	28.7 ab	58.7	159.2
Control	2.80 b	0.21	2.24	2.83	0.32	0.33	59.4 a	31.2 b	27.9 b	60.8	165.9
p-value	0.0011	0.9049	0.1458	0.7000	0.8265	0.4622	0.0234	0.0208	0.0160	0.6410	0.7727
Organic amendment × Herbicide
p-value	0.5721	0.7698	0.3887	0.2153	0.3608	0.1845	0.2752	0.1587	0.1504	0.8044	0.5912

N = nitrogen, P = phosphorus, K = potassium, Ca = calcium, Mg = magnesium, S = sulfur, B = boron, Zn = zinc, Mn = manganese, Fe = iron, Cu = copper. Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). Trees in compost-treated plots had significantly more leaf potassium (1.98%) than trees in humic acid (1.55%) and non-amended control plots (1.51%). There was a significant effect of herbicide treatment on the leaf nutrients. Trees from glyphosate and glufosinate-treated plots had significantly more leaf nitrogen (2.47% and 2.40, respectively) than control plots (2.23%) and flumioxazin-treated plots (2.26%). Leaf phosphorous content was significantly higher in glyphosate and glufosinate-treated plots (0.21%) compared to control plots (0.19%). Leaf sulfur, zinc, and manganese contents were significantly higher in glyphosate-treated plots (0.25%, 34.0 ppm, and 19.5 ppm, respectively) compared to control plots (0.22%, 30.3 ppm, and 16.7 ppm, respectively). None of the macro- and micronutrients were impacted significantly by the interaction of organic amendments and herbicides. Similar trends were found in July 2023, two years after treatment application. Trees in compost-amended plots had more leaf potassium (2.37%) and boron (61.8 ppm) than trees in humic acid plots (2.07% and 55.3 ppm, respectively) and non-amended plots (2.16% and 55.0 ppm, respectively), but the other leaf macro- and micronutrients were not affected by organic amendment treatments. Trees in glyphosate- and glufosinate-treated plots had significantly more leaf nitrogen (3.01% and 3.02%, respectively) compared to trees in the control plots (2.80%), and trees in glyphosate-treated plots had more zinc (33.9 ppm) and manganese (31.1 ppm) than trees in control plots (31.2 ppm and 27.9 ppm, respectively). Conversely, trees in glyphosate-treated plots had less boron (53.8 ppm) than trees in control plots (59.4 ppm). The interaction of treatments did not influence the leaf nutrients in July 2023.

3.4. Fruit Drop

The fruit drop measured in February 2022 was 24.7% on average and was not significantly influenced by organic amendment or herbicide treatment (

Table 3. Fruit drop, fruit yield, and fruit quality of ’Valencia’ trees grown in southwest Florida with different organic amendments and herbicides.

	Fruit Drop (%)		Fruit Yield (kg/Tree)		Juice (%)		TSS (°Brix)		TSS/TA
	2022 1	2023 2	2022	2023	2022	2023	2022	2023	2022	2023
Organic amendment
Compost	20.6	61.0 a	3.5	1.7	43.4	47.5	7.1	7.1	9.3	8.7
Humic	28.1	65.9 a	2.8	1.8	43.7	47.5	7.2	7.1	8.8	8.7
Control	25.3	71.3 a	2.8	1.4	42.5	47.1	7.00	7.1	9.2	8.9
p-value	0.2399	0.0493	0.1342	0.4566	0.1017	0.7708	0.2752	0.9337	0.4252	0.5243
Herbicide
Glyphosate	26.8	69.4 a	3.1	1.9	43.8	46.9	7.1	7.0 ab	8.9	8.9
Glufosinate	24.0	68.6 a	3.0	1.4	43.3	46.8	7.1	6.9 b	9.4	8.8
Flumioxazin	23.2	67.8 a	3.0	1.7	43.5	47.5	7.2	7.1 ab	9.1	8.7
Control	24.6	58.5 b	3.1	1.6	42.3	47.1	7.0	7.2 a	9.0	8.6
p-value	0.5396	0.0054	0.9912	0.3644	0.1430	0.1370	0.1489	0.0344	0.2932	0.6800
Organic amendment × Herbicide
p-value	0.5660	0.6180	0.8213	0.9890	0.6211	0.5343	0.4249	0.5791	0.1394	0.7873

¹ Fruit drop percentage calculated only for February 2022 at the time of harvest. ² Cumulative fruit drop percentage calculated for a 4-month observation period from November 2022 to February 2023 before the final harvest. Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). Compost and humic acid applications reduced the cumulative fruit drop measured over the 4-month period from November 2022 to February 2023 compared with the non-amended control; however, differences between means were not significant. Glyphosate, glufosinate, and flumioxazin treatments significantly increased the cumulative fruit drop (67.8–69.4%) compared to the control (58.5%). No interaction effects between organic amendments and herbicides were observed for either season.

3.5. Yield and Juice Quality

The fruit yield per tree was low overall because of the young age of the trees and weather events. In 2022, the average yield was 3.0 kg per tree, while in 2023, following Hurricane Ian, it was only 1.7 kg. There were no significant effects of organic amendment and herbicide treatments in either year (Table 3).

In 2022, there were no significant effects of organic amendments, herbicides, and their interactions on the percent juice, TSS, and the TSS/TA ratio (Table 3). In 2023, TSS was significantly influenced by the herbicide treatment. Fruits from glufosinate-treated plots had lower TSS (6.9) than fruits from the control plots (7.2). There were no significant differences for the other juice quality variables measured in that year.

3.6. Root Health

The root density ranged from 0.31 to 0.51 mg/cm³, and there were no significant effects of organic amendments, herbicides, or their interaction (Table S3).

There were no differences in the fibrous root respiration rate in response to the organic amendments, herbicides, and their interaction from April to October 2022 (

Table 4. Fibrous root respiration rates of ‘Valencia’ trees grown in southwest Florida grown with different organic amendments and herbicides.

	Respiration Rate (mg CO2 g−1h−1)
	April 2022	July 2022	October 2022	January 2023	April 2023
Organic amendment
Compost	1.80	1.72	1.03	1.31 a	1.56
Humic acid	1.57	1.14	0.92	1.10 ab	1.46
Control	1.60	1.57	0.86	1.03 b	1.50
p-values	0.6263	0.0862	0.0845	0.0130	0.8871
Herbicide
Glyphosate	1.49	1.30	0.88	1.10 ab	1.31
Glufosinate	1.48	1.49	0.99	1.16 ab	1.72
Flumioxazin	1.91	1.64	0.94	1.32 a	1.50
Control	1.80	1.48	0.93	1.02 b	1.49
p-value	0.2703	0.5136	0.6822	0.0482	0.4233
Organic amendment × Herbicide
p-value	0.4144	0.8815	0.0186	0.2004	0.9808

Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). The interaction effect was significant in October 2022, but the means separation was not significant (Table S4). In January 2023, the respiration rate was significantly affected by both organic amendment and herbicide treatments. Roots from compost-amended plots had a significantly higher root respiration rate (1.31 mg CO₂ g⁻¹h⁻¹) than roots from the non-amended control plots (1.03 mg CO₂ g⁻¹h⁻¹). Roots from the flumioxazin-treated plots had a higher root respiration rate than roots from the control plots, but there was no significant interaction of organic amendments and herbicides. In April 2023, treatments and their interactions did not influence the root respiration rate.

There were no significant effects of the treatments and their interactions on the first-, second-, or third-order SRL from April to October 2022 (Table S5). In January 2023 and April 2023, roots from glyphosate-treated plots had a higher first-order SRL (28.1 m/g) than roots from control plots (24.7 m/g), but there was no effect from the organic amendments (

Table 5. Specific root lengths (SRLs) of 1st-, 2nd-, and 3rd-order roots of ‘Valencia’ trees grown in southwest Florida with different organic amendments and herbicides.

	1st Order SRL (m/g)		2nd Order SRL (m/g)		3rd Order SRL (m/g)
	January 2023	April 2023	January 2023	April 2023	January 2023	April 2023
Organic amendment
Compost	26.1	25.6	19.2 a	17.1 ab	11.6	10.8 a
Humic acid	27.6	26.8	20.6 a	19.5 a	11.8	10.7 a
Control	25.7	23.8	18.2 a	16.6 b	11.5	9.2 b
p-value	0.2503	0.1164	0.0404	0.0075	0.9592	0.0159
Herbicide
Glyphosate	28.1 a	27.3 a	20.9	19.7	12.6	10.9
Glufosinate	27.0 ab	25.2 ab	19.5	17.0	12.0	10.3
Flumioxazin	26.0 ab	25.2 ab	19.1	17.5	10.9	9.7
Control	24.7 b	23.9 b	18.0	16.8	10.9	9.9
p-value	0.0056	0.0305	0.0828	0.0501	0.4444	0.2888
Organic amendment × Herbicide
p-value	0.2151	0.0841	0.3662	0.0989	0.6534	0.0035

Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). There was a significant effect of the organic amendment on the second-order SRL, but the mean separation was only significant in April 2023 when roots from humic acid-amended plots had a higher second-order SRL (19.5 m/g) than roots from the non-amended control plots (16.6 m/g). There was no significant effect of any of the treatments for the third-order SRL in January 2023, but there was a significant interaction between the organic amendments and herbicides in April 2023. Roots from non-amended × glyphosate-treated plots had a higher third-order SRL (12.5 m/g) than roots from non-amended × flumioxazin-treated plots, non-amended × herbicide control plots, and non-amended × glufosinate treated plots (8.2–8.4 m/g) (Table S6). Overall, compost- and humic acid-amended plots had a higher third order SRL than non-amended plots in April 2023.

3.7. Soil Properties

The soil volumetric water content (VWC) was measured monthly. The VWC was significantly affected by organic amendment and month, but not by their interaction (Figure 3). Compost-amended plots had a significantly higher VWC (6.9% on average) than humic acid-amended plots (5.6% on average) and non-amended plots (4.4% on average). The highest VWC was observed in June 2022 (16.5%), while the lowest was observed in December 2022 (2.5%) and March 2023 (1.8%).

Soil nutrients were measured in July of each year. The organic amendment treatments did not induce any significant differences for the soil nutrients in 2022, one year after treatment application, except for boron, zinc, and copper (

Table 6. Soil nutrient content of trial in southwest Florida under different organic amendments and herbicides.

	P (ppm)	K (ppm)	Mg (ppm)	Ca (ppm)	S (ppm)	B (ppm)	Zn (ppm)	Mn (ppm)	Fe (ppm)	Cu (ppm)
July 2022
Organic amendment
Compost	63.4	18.9	60.1	634.1	6.1	0.28 a	5.6 a	7.4	44.5	26.9 b
Humic acid	67.5	13.4	53.4	525.2	6.8	0.23 a	6.0 a	7.9	48.8	33.4 ab
Control	63.9	13.1	51.4	538.7	6.7	0.24 a	6.8 a	8.9	49.9	34.6 a
p-value	0.8606	0.0503	0.3509	0.0801	0.6299	0.0351	0.0258	0.0518	0.1436	0.0136
Herbicide
Glyphosate	54.6	11.8 b	43.9 b	502.7	5.1	0.23	6.1	8.1	46.5	33.4
Glufosinate	69.0	15.1 ab	55.0 ab	561.3	6.8	0.27	6.2	8.5	47.7	31.6
Flumioxazin	61.5	16.6 ab	58.0 ab	575.8	6.7	0.24	6.1	7.4	45.9	30.9
Control	76.5	17.0 a	62.8 a	624.3	7.6	0.25	6.2	8.2	50.7	30.7
p-value	0.1201	0.0297	0.0130	0.0735	0.1201	0.2981	0.9935	0.4193	0.2026	0.5115
Organic amendment × Herbicide
p-value	0.7606	0.9241	0.8022	0.5264	0.1347	0.4032	0.4104	0.1347	0.2251	0.3559
July 2023
Organic amendment
Compost	113.8 a	47.4 a	119.8 a	1291.9 a	17.6 a	0.49 a	8.23	11.1	53.2	33.3 b
Humic acid	90.2 a	28.6 b	71.0 b	565.4 b	12.5 b	0.24 b	7.71	10.0	49.8	39.0 ab
Control	81.0 a	25.2 b	69.8 b	567.6 b	10.9 b	0.24 b	7.21	10.4	53.6	41.8 a
p-values	0.0218	<0.0000	<0.0000	<0.0000	<0.0000	<0.0000	0.3642	0.3636	0.4351	0.0108
Herbicide
Glyphosate	74.3 b	26.2 b	74.5	733.3	11.7	0.28	8.08	10.6	52.9	38.5
Glufosinate	91.6 ab	33.0 ab	81.1	774.3	13.8	0.29	7.94	10.5	51.9	40.5
Flumioxazin	99.2 ab	34.2 ab	91.6	830.9	14.5	0.31	7.48	10.0	50.9	36.9
Control	114.8 a	41.4 a	100.3	894.8	14.6	0.35	7.36	10.9	53.1	36.0
p-value	0.0364	<0.0000	0.0910	0.2828	0.1924	0.1831	0.5640	0.7840	0.7891	0.2160
Organic amendment × Herbicide
p-value	0.8595	0.5373	0.7884	0.8747	0.9734	0.9870	0.8107	0.2141	0.4303	0.1859

Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). While the mean separation was non-significant for boron and zinc, compost-amended plots had significantly less copper (26.9 ppm) than non-amended control plots (34.6 ppm). Among herbicide treatments, glyphosate-treated plots were lower in potassium (11.8) and magnesium (43.9 ppm) than in control plots (17.0 ppm and 62.8 ppm, respectively). In 2023, two years after treatment application, organic amendments were seen to have significantly influenced most of the soil nutrients except zinc, manganese, and iron. Compost-amended plots had more potassium (47.4 ppm), magnesium (119.8 ppm), calcium (1291.9 ppm), sulfur (17.6 ppm), and boron (0.49 ppm) than humic acid- and non-amended plots. Conversely, as in the previous year, more copper was found in the non-amended control plots (41.8 ppm) compared to the compost-amended plots (33.3 ppm). Although the mean separation was not significant, compost-treated plots tended to have more phosphorous than the other plots. Herbicide treatment affected the soil phosphorus and potassium contents, which, as in the previous year, were lower in the glyphosate-treated plots (74.3 ppm and 26.2 ppm, respectively) than the control plots (114.8 ppm and 41.4 ppm, respectively). There was no significant interaction between organic amendments and herbicides for the soil nutrient content. Most soil nutrients were generally higher in content in 2023 compared to 2022.

In July 2022, organic amendments and herbicides did not affect soil organic matter, cation exchange capacity (CEC), pH, nitrate, or ammonium content significantly (

Table 7. Soil physicochemical properties of trial in southwest Florida under different organic amendments and herbicides.

	OM (%)	CEC	pH	NO3− (ppm)	NH4+ (ppm)
July 2022
Organic amendment
Compost	0.57	4.4	6.7	0.52	1.05
Humic acid	0.53	3.9	6.3	0.57	0.95
Control	0.57	4.1	6.4	0.51	1.10
p-value	0.8252	0.1915	0.0545	0.9091	0.4887
Herbicide
Glyphosate	0.49	3.5 a	6.4	0.48	0.98
Glufosinate	0.55	4.2 ab	6.5	0.52	1.05
Flumioxazin	0.57	4.2 ab	6.6	0.60	0.94
Control	0.62	4.7 b	6.6	0.55	1.16
p-value	0.0788	0.0117	0.4514	0.7891	0.4380
Organic amendment × Herbicide
p-value	0.8445	0.8541	0.8318	0.2929	0.1755
July 2023
Organic amendment
Compost	1.11 a	8.3 a	7.2 a	2.6 a	0.20
Humic acid	0.79 b	4.8 b	6.3 b	1.5 b	0.09
Control	0.73 b	4.7 b	6.2 b	1.2 b	0.03
p-value	<0.0000	<0.0000	<0.0000	0.0008	0.3529
Herbicide
Glyphosate	0.74	5.2	6.5	1.3	0.04
Glufosinate	0.84	5.7	6.5	1.8	0.06
Flumioxazin	0.94	6.2	6.5	2.2	0.04
Control	0.98	6.6	6.7	1.9	0.30
p-values	0.0760	0.0617	0.6186	0.3024	0.0674
Organic amendment × Herbicide
p-value	0.7805	0.9293	0.8477	0.8122	0.1068

OM = organic matter, CEC = cation exchange capacity. Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). However, glyphosate-treated plots had a lower CEC (3.47 meq/100 g) compared to the control plots (4.67 meq/100 g). The interaction of organic amendments and herbicides did not influence these soil properties. In July 2023, compost-amended plots had a significantly higher soil organic matter content (1.11%), CEC (8.3 meq/100 g), pH (7.2) and nitrate content (2.64 ppm) than humic acid- and non-amended plots (0.73/0.79%, 4.7/4.8, 6.2/6.3, and 1.18/1.45 respectively). There was no significant interaction of organic amendments and herbicides for any of these soil properties. Soil organic matter, CEC, and NO₃ were overall higher in 2023 compared to 2022.

3.8. Weed Biomass and Diversity

The weed biomass was assessed in March 2022, August 2022, and April 2023. There were no significant effects of the organic amendments on the weed biomass at any of the time points, but there was a significant herbicide effect (

Table 8. Weed biomass of trial in southwest Florida soil under different organic amendments and herbicides.

	Weed Biomass (g)
	March (2022)	August (2022)	April (2023)
Organic amendment
Compost	42.7	108.0	16.78
Humic	45.3	91.5	23.95
Control	32.7	92.6	24.43
p-value	0.3303	0.2873	0.2901
Herbicide
Glyphosate	7.3 c	59.2 c	10.0 c
Glufosinate	32.7 b	85.9 b	10.7 c
Flumioxazin	56.9 a	118.2 a	26.3 b
Control	63.9 a	126.0 a	39.9 a
p-value	<0.0000	<0.0000	<0.0000
Organic amendment × Herbicide
p-value	0.1785	0.8808	0.0339

Different letters indicate significant differences (p < 0.05) within columns according to Tukey’s honestly significant difference test.

). At all sampling times, the post-emergent herbicides glyphosate and glufosinate controlled weeds better than the pre-emergence flumioxazin. Glyphosate-treated plots had the lowest weed biomass in March 2022 and August 2022 (7.3 g and 59.2 g, respectively), followed by glufosinate-treated plots (32.7 g and 85.9 g, respectively). There was no significant difference in the weed biomass between flumioxazin-treated and control plots. In April 2023, there was a significant interaction of organic amendments and herbicides, with the lowest weed biomass found in humic acid-amended × glyphosate-treated plots (3.9 g) and the highest in humic acid-amended × herbicide control plots (56.7 g) (Table S7).

The weed diversity (species composition) was assessed in June 2022 and January 2023. The predominant weed species were Spanish needle (Bidens alba), torpedo grass (Panicum repens), Bermuda grass (Cynodon dactylon), and pusley (Richardia brasiliensis). In June 2022, less torpedo grass (34%) and more Spanish needle (41%) were found in compost-amended plots compared to humic-acid amended plots (54% and 23%, respectively) and control plots (68% and 24%, respectively), and more Bermuda grass was found in compost- and humic acid-amended plots (23% and 19%, respectively) compared to the non-amended control plots (4%) (Figure 4). In January 2023, more pusley and Spanish needle were observed in the control plots (7% and 52%, respectively) than in humic acid- (0% and 36%, respectively) and compost-amended plots (1% and 39%, respectively), whereas more torpedo grass was observed in compost-amended plots (52%) compared to humic acid-amended (34%) and control plots (25%).

The post-emergence herbicides glufosinate and glyphosate eliminated torpedo grass populations in June 2022, while selecting Spanish needle (56% and 64%, respectively) and Bermuda grass (44% and 36%, respectively) (Figure 5). Flumioxazin-treated and control plots contained 63% and 47%, respectively, of torpedo grass, 29% and 27%, respectively, of Spanish needle, and 6% and 22%, respectively, of Bermuda grass. In January 2023, other weed species such as cutleaf evening primrose (Oenothera laciniata) and balsam apple (Momordica balsamina) were observed across the experimental site, indicating a shift in the weed population. In glyphosate-treated plots, the abundance of other weeds was 50%, while it was 15–20% in the other plots. No torpedo grass was observed in glyphosate-treated plots in contrast to the other plots where it was abundant (32–37%). Spanish needle was more abundant in glyphosate- and flumioxazin-treated plots (50% and 52%, respectively) compared with glufosinate-treated plots and the control (39% and 36%, respectively).

Although glyphosate suppressed weeds effectively, it resulted in a higher occurrence of broadleaf weeds in August 2022 (92.9%) and April 2023 (80.0%), whereas glufosinate-treated plots tended to produce relatively more grass weeds (58.2% and 68.6%) compared to the other treatments (Figure 6).

4. Discussion

Organic amendments did not influence young tree growth over the experimental period, suggesting that more time may be needed to see any major changes in tree performance. In contrast to the organic amendments, herbicide treatment affected the tree growth rate during the experimental period. Both post-emergent herbicides, glyphosate and glufosinate, increased the growth, likely due to reduced resource competition with weeds.

While organic amendments did not significantly affect the weed biomass, the use of glyphosate and glufosinate provided better weed control than flumioxazin throughout the experiment. Flumioxazin is a pre-emergence herbicide, which prevents the emergence of new weeds but does not kill existing weeds. The lack of efficacy when used alone suggests that flumioxazin needs to be combined with post-emergence herbicides for efficient weed suppression, especially at locations with high weed pressure such as in this study. The weed biomass at the study site was double in late summer (August) compared to spring (April) because of frequent rainfall events typical for Florida during the summer season. After two years of repeated application, glufosinate provided similar weed suppression as glyphosate, which suggests that it may be used as a glyphosate alternative. Glufosinate treatments provided the better control of broadleaf weeds, while glyphosate was more effective at controlling grass weeds, and it allowed the establishment of broadleaf weeds like Spanish needle. Takano and Dayan [34] reported that broad leaf weeds are more susceptible to glufosinate, which aligns with our findings. Repeated herbicide applications resulted in a gradual change in weed species composition, although seasonal differences may have contributed to these changes. Existing species like Guinea grass and Bermuda grass and new weeds like primrose and balsam apple rapidly established in the experimental plots, taking over previous dominant species like torpedo grass and Spanish needle near the end of the experiment.

Despite not having a significant impact on tree growth, the organic amendments impacted several soil properties. Both compost and humic acid increased the soil volumetric water content (VWC) throughout the study, although the effect was more moderate in the humic acid-treated soils. The water-holding capacity of Florida’s sandy soils is directly related to their organic matter content, which in turn can be increased through the application of organic amendments [44]. In this study, we observed significant increases in total soil organic matter with the use of compost, but not until two years of repeated application. A four-year study conducted in a sandy soil environment in British Columbia also did not find an increase in organic matter until two years of organic amendment applications, and a higher water retention capacity was observed only after three years [45]. A three-fold increase in organic matter content was found after nine years of organic amendment application in a previous study in Florida [46]. The improved soil water-holding capacity observed in our study suggests that compost may help reduce the irrigation frequency in Florida’s citrus production systems. Despite the higher soil moisture content induced by the organic amendments, no significant increase in weed pressure was observed during the experimental period, likely because of the relatively larger influence of the rainfall in this region.

The increased soil organic matter in compost-amended plots also impacted other soil properties, such as CEC and pH, which increased, particularly after two years of application. The optimal soil pH range for citrus is 6–6.5, but was increased to 7.2 in 2023 in response to the compost applications. This must be considered when implementing the use of compost for citrus management, as a pH over 7 decreases the availability of potassium, calcium, manganese, zinc, and iron [47]. In contrast, a higher CEC results in a larger number of exchangeable cations retained in the soil for plant uptake [48]. In accordance, the compost-amended soil also had a higher nutrient content (P, Mg, Ca, S, and B) after two years of application. In contrast, compost applications consistently reduced the soil copper content. Carboxyl groups present in the compost may have chelated some of the copper ions to form soluble complexes, which may then have leached to deeper zones in the soil. The increased soil moisture in response to compost applications may have also enhanced this process [49].

The soil nutrient content was also affected by the herbicide treatments. Glyphosate reduced the soil potassium, magnesium, and phosphorous content. However, this did not result in concurrent leaf nutrient reductions. Glyphosate is a soluble organic acid that has been reported to chelate soil nutrients, especially divalent cations such as magnesium [50]. Glyphosate is also known to compete with the phosphate group for adsorption sites in the soil [51], increasing nutrient leaching. Both effects may have contributed to the reduction in some of the soil nutrients observed in our study. Reduced competition with weeds and the resulting increase in plant growth and nutrient absorption may also have contributed to this effect.

Both organic amendments and herbicide treatments significantly influenced the leaf nutrient content. The compost appeared to provide additional nutrients to the trees, as soil potassium and boron content were increased in soil and leaves in both years. Trees from plots treated with herbicides had greater concentrations of some nutrients, likely due to reduced competition with weeds. For example, trees from plots treated with glyphosate and glufosinate consistently contained more leaf nitrogen compared to the untreated controls. Glyphosate also increased the levels of some of the other nutrients (N, P, S, Zn, and Mn). Kaur et al. [52] defined weeds as ‘luxury consumers’ who rapidly exploit the nutrients and other resources applied to soil. In Florida, especially during the rainy season, weeds grow fast and can quickly overtake the space underneath the tree canopy [32]. Previous studies on other tree crops also demonstrated an improved tree nutrient status when weeds were controlled effectively [53].

The respiration rate of the fibrous roots indicates their metabolic state, and a higher root respiration rate generally represents higher metabolic activity, which supports plant growth and maintenance function [54]. In our study, the fibrous root respiration rate appeared to have seasonal variations, but was also impacted by the treatments. There was a trend for higher root respiration rates in response to compost treatments, but effects were only significant in January 2023 compared to the control. Seasonal fluctuation in root respiration can be related to soil temperature and moisture [53]. A study with olive trees under different moisture regimes found a higher fibrous root respiration rate with increasing water availability [55]. Similarly, in grapefruit trees, the root respiration rate decreased with the depletion in soil moisture [56]. Root respiration is also known to correlate with nitrogen availability and root nitrogen concentration [57]. In our study, after two years of compost application, there was a significantly greater content of several nutrients, including nitrate, which along with the higher soil moisture content may have contributed to the higher root respiration rate.

The specific root length (SRL) explains the metabolic costs associated with root elongation and varies with the root order, where a higher SRL indicates less energy/carbon required for the production and maintenance of roots [58]. Resource acquisition depends more on root length than root biomass [59], and a higher SRL may allow HLB-affected citrus trees to compensate for disease-induced root decline. Consistent with other findings [60], in our study, the SRL decreased with increasing root order. Lower-order roots are responsible for resource exploration and acquisition, while higher-order roots are involved in transportation, storage, and lateral root production [61]. While the first-order SRL was not affected by the organic amendments during the experimental period, glyphosate induced a higher SRL in year 2 of the study. Previous studies have suggested that glyphosate, when applied at a high rate, may accumulate in the root zone and negatively affect root growth [62]. The increase in the first-order SRL induced by glyphosate suggests that this was not the case in our study. First-order roots are constantly reconstructed, a process that is affected by nutrient and water availability, available space, competition, and other factors [61,63]. The reduced competition from weeds because of glyphosate application may be the main reason for the observed higher SRL. Herbicide treatment did not affect the SRL of the other, higher-order roots, but organic amendments did after two years of study. Improved soil health caused by the organic amendments over time may have contributed to the increased SRL of these roots.

The fruit yield was generally low in both harvest seasons because of the tree’s young age and the hurricane in 2022. However, despite some of the positive effects on plant growth, root properties, and soil health, neither organic amendment nor herbicide treatments induced any significant yield effects, nor did they affect the juice quality. Safaei Khorram et al. [64] also did not find any influence of organic amendments on the fruit yield of apple trees over a four-year application period. More time is likely needed for tree crops to show any benefits derived from organic amendments on growth and productivity. Despite the lack of an effect on yield, herbicide applications appeared to have increased the pre-harvest fruit drop, suggesting care must be taken when applying herbicides to reduce any harmful effects on crop production. The tendency of glyphosate to enhance premature fruit drop was previously reported [65].

CLas titer levels (Ct values) were similar across all treatments in both leaves and roots, suggesting no direct effect on the pathogen population. These results differ from those of a previous study in Florida, where compost applied at 30 tons/hectare (12 tons/acre) reduced roots CLas titer levels in trees [66]. The lower compost rate used in this study along with other differences may have contributed to the different results.

5. Conclusions

Soil health is a function of the soil’s physical (texture, porosity, VWC, etc.), chemical (pH, CEC, organic matter, nutrient content, etc.), and biological (rhizosphere microbial activity) properties. In our study, compost application positively influenced some of these properties, improving the soil environment for citrus tree growth. Compost application also improved nutrient uptake and root physiological functions like fibrous root respiration, although the effects were not apparent until repeated applications were performed. Despite these positive effects, organic amendments did not affect tree growth, fruit yield, or fruit quality during the study period, although unfavorable weather events may have influenced the results. Nevertheless, the observed positive effects, albeit subtle, suggest that growers may benefit from longer-term applications of compost. Longer-term studies in different locations need to be conducted to assess whether effects are cumulative and result in more substantial benefits to tree growth and productivity. The post-emergence herbicides glyphosate and glufosinate effectively managed weeds, improved nutrient uptake, and improved tree growth significantly. The judicious application of herbicides combined with compost application may be a rational cultural practice to maintain citrus production in the HLB era, although effects may not be evident until several years later.