Bioassay experiments with plant parameters including shoot dry mass in Expt. 2 (A) and Expt. 3 (B) and soil–plant analysis development (SPAD) chlorophyll index for Expt. 2 (C) and Expt. 3 (D). The letters indicate mean comparison for the interaction of species and applied heavy metal concentration using Tukey’s honestly significant difference at α = 0.05.
Visual symptoms of heavy metal toxicity in mustard, which were only observed with the 30- and 60-mg/kg Cd and Pb treatments in high-dose bioassay (Expt. 2).
Greenhouse Expt. 1 with Cd and Pb concentration in shoot tissue (A) and proportion of applied Cd and Pb taken up into shoot tissue per plant [B, heavy metal (HM) shoot concentration × shoot dry mass/HM applied per plant]. In comparison with the measured concentrations in (A), the Florida regulatory thresholds are 0.5 mg/kg for Pb and 0.2 mg/kg for Cd. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05 by heavy metal.
Uptake of heavy metal (HM) into shoot tissue in the high-dose bioassay (Expt. 2). (A and B) Average Cd (A) or Pb (B) tissue concentration by HM treatment level and species. (C and D) Proportion of applied Cd (C) or Pb (D) taken up into shoot tissue (HM shoot concentration × shoot dry mass/HM applied per plant). Note that the scale of the y axis varies between graphs. The dashed lines on the tissue concentration graph (B) indicates the Florida regulatory threshold for Pb of 0.5 mg/kg, and the threshold for Cd in graph (A) is 0.2 mg/kg. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05.
Uptake of heavy metal (HM) into shoot tissue in the low-dose bioassay (Expt. 3). (A to C) Average As (A), Cd (B), and Pb (C) concentrations in shoot tissue by HM treatment level and species. (D to F) The proportion of applied As (D), Cd (E), and Pb (F) taken up into shoot tissue by HM treatment level and species. The dashed lines on the tissue concentration graphs (A) and (C) indicate the Florida regulatory threshold for As and Pb, and the threshold for Cd is 0.2 mg/kg. Note that the scales of the y axes vary between graphs. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05.
Click on author name to view affiliation information
Medicinal and food crops including hemp (Cannabis sativa L.) are routinely tested for heavy metal (HM) contaminants including As, Cd, and Pb per regulatory guidelines. Heavy metals have been found in greenhouse substrates, including peat and coconut coir, and are commonly used in hemp production, which underscores the need to evaluate the risk of HM contamination of hemp from these production inputs. The objective was to quantify uptake of As, Cd, and Pb in coconut coir substrate into shoot tissue using bioassays and compare heavy metal uptake between brassicas and hemp. In Expt. 1, ‘Maverick’ hemp was grown for 8 weeks in a greenhouse in coconut coir–filled containers that were dosed before planting with 0.0, 0.5, 1.0, 4.8, or 9.7 mg of soluble Cd and Pb per kg of dried coir (mg·kg−1). ‘Maverick’ hemp, ‘Savanna’ mustard (Brassica juncea L.), and ‘Red Russian’ kale (Brassica oleracea L.) seedlings were also grown in a growth chamber for 4 weeks in propagation trays with HM application levels of 0, 3, 15, 30, and 60 mg·kg−1 of Cd and Pb (Expt. 2) or 0.0, 0.5, 1.0, 3.0, and 5.0 mg·kg−1 of As, Cd, and Pb (Expt. 3). In all experiments, Cd tissue concentration and mg of Cd per plant increased as HM levels in the substrate increased. Cd exhibited the greatest plant uptake, followed by As and then Pb. Average As and Cd concentrations in hemp tissue were above a typical regulatory threshold for medicinal cannabis of 0.2 mg·kg−1 of dried tissue in all experiments when HM was applied at 0.5 mg·kg−1 of substrate or greater. In contrast, Pb was the least plant-available HM, with shoot tissue concentrations in hemp below the regulatory threshold of 0.5 mg·kg−1 except when substrate Pb reached 30 or 60 mg·kg−1. The results therefore indicate a higher potential risk of contamination from As and Cd for medicinal hemp than Pb at a given HM concentration. Mustard had the greatest accumulation of As and Cd, followed by kale and hemp, emphasizing that HM accumulation is also of concern for controlled environment production of brassica crops.
Heavy metal (HM) contamination is a significant issue in medicinal hemp and hemp products (Dryburgh et al. 2018; Holt et al. 2023; Wakshlag et al. 2020; Wedman-St. Louis 2018). Heavy metals including As, Cd, and Pb pose a serious threat to human health and safety (Wieliczko 2020). Each of these HMs have been identified as carcinogens and have been linked to other serious health issues such as reduced fertility and intelligence quotient (Guo et al. 2020; López-Botella et al. 2021). In the United States, allowable thresholds for these metals are low and can vary by state (Peng and Shahidi 2021). In Florida, regulatory thresholds for As, Cd, and Pb in medicinal hemp flower are 0.2, 0.2, and 0.5 mg·kg−1 of dried tissue, respectively (Florida Department of Health 2020). One study found that mean concentrations of As, Cd, and Pb for 28 medicinal hemp samples were above these regulatory thresholds (Amendola et al. 2021), and in another study, at least half of hemp products tested contained detectable levels of Pb (Seltenrich 2019).
When an analytical report indicates HM contamination, hemp producers often test production inputs such as substrate, fertilizer, and irrigation water to determine the contaminant source of origin. Inputs such as irrigation water or fertilizer pose a high risk of contaminating a hemp crop, because they may be applied directly to plant surfaces over the course of a crop cycle and are repeatedly applied. However, HM have also been found in substrates such as peat and coconut coir at levels ranging from 0.01 to 16 mg·kg−1 Cd and 0.74 to 800 mg·kg−1 Pb (Abreu et al. 2005; Fritz and Venter 1988; Livett et al. 1979; Smieja-Krol et al. 2010). Substrates amended with compost can exhibit high levels of HM, with one study finding up to 386 mg·kg−1 Pb in composted material (Vukobratović et al. 2013). Field soil also has the potential to contain high levels of HM, although the concentration varies greatly between locations (Radanovic et al. 2001). Despite the potentially high HM levels found in substrates, it is possible that only a fraction of these metals are available for plant uptake (McLaughlin et al. 2000a, 2000b).
The current industry standard for measuring heavy metals in both hemp products and production inputs is to use inductively coupled plasma (ICP)–mass spectrometry (MS) or ICP–optical emission spectrometry (OES), because of their ability to independently quantify elements at trace levels with high resolution (Ammann 2007). However, before production inputs such as substrate are analyzed for HM content, they are first digested in strong acids and ionized at high temperatures before the concentration of an individual element is determined (Chen and Ma 2001). Acid digestion of the substrate is intended to quantify the complete HM content but may overestimate the plant-available fraction that would be available for plant uptake under normal environmental conditions (Maiz et al. 1997; McLaughlin et al. 2000b; Zimmerman and Weindorf 2010). Alternative extraction methods can be used to quantify the plant-available fraction of HM, including a variety of laboratory procedures such as diethylenetriaminepentaacetic acid extractions or sequential extraction procedures and bioassays with HM-accumulating plant species (McLaughlin et al. 2000b).
Bioassays are a method of detecting, measuring, and assessing the risk of a contaminant or substance of interest by using living organisms as the extractant or measurement tool (Christofi 2005). Bioassays are considered well-suited for measuring metal availability in the rhizosphere, because they integrate all aspects of soil–plant interactions and environmental factors that influence HM uptake (McLaughlin et al. 2000b; Nazir et al. 2011). Species selection for the bioassay method is an important consideration, because species and varieties differ in their ability to accumulate HM (Husain et al. 2019; Rubio et al. 1994). Hemp is a known hyperaccumulating plant species of HM and has been studied extensively for its phytoremediation potential (Ćaćić et al. 2019; Citterio et al. 2003). Mustard and kale have also been documented as hyperaccumulators of HM, and several studies have been conducted to examine their phytoremediation potential (Mourato et al. 2015; Qadir et al. 2004; Radulescu et al. 2013). Because of their documented capacity to hyperaccumulate HM, Cannabis sativa, Brassica juncea, and Brassica oleracea were chosen as model crops for our bioassay experiments.
Although substantial research has examined the potential risks of HM contamination in field soils, there has been limited evaluation of the risk of HM contamination in soilless substrates such as coconut coir. Given the limitations of total digestion for assessing plant-available HM in substrate, the low regulatory thresholds for hemp growers, and the popularity of soilless substrate for hemp production, there is a need to evaluate the availability of HM in greenhouse substrates such as coconut coir.
The objectives were to quantify plant availability of As, Cd, and Pb in coconut coir substrate using bioassays and to compare HM uptake between hemp and two brassica crop species, mustard and kale. It was hypothesized that not all HM applied to the coconut coir substrate would be readily absorbed and translocated to shoots, because of barriers such as precipitation or adsorption to the substrate matrix.
Three plant experiments were conducted, with HM incorporated into substrate before planting. Expt. 1 represented a crop production scenario from planting through to the flowering stage over 8 weeks. ‘Maverick’ hemp seedlings were transplanted into containers filled with coconut coir that had received applications of 0 to 10 mg of Cd and Pb per kg of dried substrate before planting. The plants were fertigated with each irrigation at 200 mg·L−1 N from 15N–2.2P–12.5K water-soluble fertilizer with micronutrients (Peter’s Excel 15–5–15 Cal-Mag Special; ICL Fertilizers, Summerville, SC, USA) for 5 weeks, and fertigation was increased to 300 mg·L−1 N for the final 3 weeks. The water source was tap water [pH 7.5, electrical conductivity (EC) 0.3 mS/cm, alkalinity 45 mg·L−1 CaCO3]. The plants were drip-irrigated when moisture level dried to ∼50% of container capacity based on checking gravimetrically. A saucer was placed under each container to allow reabsorption of any leachate.
Expts. 2 and 3 were short-term 4-week bioassays more suitable for application as a quality control testing procedure by a substrates laboratory. ‘Maverick’ hemp, ‘Savannah’ mustard, and ‘Red Russian’ kale seedlings were grown in a growth chamber in propagation trays with 0.0 to 60.0 mg of Cd and Pb per kg of dried coir in Expt. 2, and 0.0 to 5.0 mg of As, Cd, and Pb in Expt. 3. The plants were initially fertigated with each irrigation at 75 mg·L−1 N from 15N–2.2P–12.5K water-soluble fertilizer with micronutrients (Peter’s Excel 15–5–15 Cal-Mag Special; ICL Fertilizers) and then at 100 mg·L−1 N at day 10 (Expt. 2) and 150 mg·L−1 N at day 15 (Expt. 3) using subirrigation when moisture level dried to ∼50% of container capacity based on checking gravimetrically, with a collection tray placed under each container to allow reabsorption of any leachate.
Potential contamination with As, Cd, and Pb was tested for production inputs. Irrigation tap water (used in Expt. 1) was combined with 15N–2.2P–12.5K water-soluble fertilizer with micronutrients (Peter’s Excel 15–5–15 Cal-Mag Special; ICL Fertilizers) at 100 mg·L−1 N for analysis, and the deionized water source (used in Expts. 2 and 3) was analyzed without fertilizer. Undiluted 16-mL samples were analyzed by Quality Analytical Laboratories (Panama City, FL, USA) using U.S. Environmental Protection Agency (EPA) method 6010B (Environmental Sampling and Analytical Methods Program 1996a) with ICP-OES (iCAP 7400 Radial; Thermo Scientific, Waltham, MA, USA). All HM samples were below the practical quantification limit (PQL) of 0.1 mg·kg−1 for As, Cd, and Pb. The same coconut coir substrate was used in all experiments (FibreDust; Sun Gro Horticulture, Agawam, MA, USA). Three 0.25-g samples of the coir were tested by ModernCanna Laboratories (Lakeland, FL, USA), using EPA microwave-assisted digestion method 3051A (Environmental Sampling and Analytical Methods Program 2007) and analysis by ICP-MS (Agilent 7800; Agilent, Santa Clara, CA, USA) using EPA method 6020A (Environmental Sampling and Analytical Methods Program 1998) with PQLs of 0.1, 0.05, and 0.1 mg·kg−1 for As, Cd, and Pb, respectively. As and Cd were below the PQL, and Pb averaged 0.21 mg·kg−1.
Expt. 1 was a randomized block design, where the main factor was HM concentration, and both Cd and Pb were applied into the same nursery plant containers. Four experimental blocks (greenhouse benches) were used with two replicate containers per HM level per block, which yielded a total of eight replicates per HM level. Five levels of HM were applied 1 week before transplanting of seedling plugs to 2.05-L containers (1-gallon “trade” nursery pots; BWI, Nash, TX, USA) filled with coconut coir. The range of HM chosen for this experiment was designed to elicit a dose–response curve and to simulate conditions that were reported in previous studies (Abreu et al. 2005; Fritz and Venter 1988; Radanovic et al. 2001). Heavy metal levels included 0.00, 0.48, 0.97, 4.84, and 9.68 mg each of Cd and Pb per kg of dry substrate, which represented 0, 0.062, 0.12, 0.62, and 1.23 mg/plant, with one plant/2.05-L container.
Expts. 2 and 3 differed from Expt. 1 in several ways to refine the bioassay method. Changes included the application method of HM to the substrate, the addition of As in Expt. 3, the range in HM dose, the inclusion of kale and mustard as model hyperaccumulating crops for comparison with hemp, and plant growth in an air-conditioned growth chamber under sole-source LED lighting. In all three experiments, only the aboveground tissue [consisting of combined flowers (only present in Expt. 3), stems, and leaves] was harvested and analyzed for HM content, because these are the portions of the plant that have economic value, and it was difficult for roots to be completely separated from substrate, or to differentiate HM on the outer surface or inside roots.
Expts. 2 and 3 used a randomized complete block design in which the main factors were HM concentration and plant species. Five HM treatments were included in Expt. 2 (0, 3, 15, 30, and 60 mg of Cd and Pb per kg of dried coir). Treatment levels in Expt. 3 had lower concentrations and included arsenic (0.0, 0.5, 1.0, 3.0, and 5.0 mg of As, Cd, and Pb per kg of dried coir). The range of HM levels chosen for all three experiments was based on previous literature examining HM concentrations in peat and soils incorporated with waste composts and was selected to mimic conditions that growers might encounter in a production setting (Abreu et al. 2005; Fritz and Venter 1988; Radanovic et al. 2001). Three plant species were grown from seed in both experiments, which included ‘Red Russian’ kale, ‘Savanna’ mustard, and ‘Maverick’ hemp. The main plot was subdivided into three blocks, each with two replicates per species per treatment, for a total of 90 replicates. Replicates were in 26 × 25.5-cm propagation trays containing 36 cells (one plant/cell) and with each tray filled with 1.55 L of 100% coconut coir.
The same source of coconut coir (FibreDust; Sun Gro Horticulture) was chosen for all three experiments. Coir is commonly used in hemp production (Zheng 2021) and could be uniformly hydrated with HM-contaminated solutions. In Expt. 1, a 123-mg·L−1 stock solution of Cd and Pb was prepared from reagent grade CdCl2 (Alfa Aesar, Haverhill, MA, USA) and PbCl2 (Acros Organics, Geel, Belgium) in deionized water. The stock solution was then pipetted into nursery containers filled with coconut coir substrate, according to their respective HM treatment levels. Nursery containers were subsequently top-irrigated to container capacity to distribute the HM throughout the substrate, and saucers were left under the containers for the duration of the experiment to ensure HMs were not lost via leaching. In all three experiments, the coir was left to equilibrate for 1 week before planting.
To achieve a higher degree of HM uniformity throughout the profile of the substrate for Expts. 2 and 3, oven-dried bricks of coconut coir were weighed and then hydrated using deionized water that had been dosed with varying concentrations of HM. Aqueous stock solutions of As, Cd, and Pb each at 1000 mg·L−1 were prepared from National Institute of Standards and Technology standard solutions (lot numbers 1359630, 130116, and 101026 respectively) and the varying quantities of HM solutions were added to volumetric flasks. Flasks were brought to 5 L of total volume before being applied to the dehydrated bricks of coir. One flask each was used for the respective As, Cd, and Pb stock solutions, and flasks were triple-rinsed with deionized water between every use to prevent cross-contamination between treatment levels. The coconut coir substrate was mechanically mixed after hydration and then daily over a 1-week period to improve homogeneity. In all three experiments, substrate pH and EC were recorded (saturated paste extraction in the first experiment, and plug-squeeze method in Expts. 2 and 3), and the results are listed in Table 1.
For Expts. 1 and 2, plant tissue and substrate samples were analyzed by Quality Analytical Laboratories with samples digested using open-vessel EPA method 3050B (Environmental Sampling and Analytical Methods Program 1996b). The 5-g samples were diluted by a factor of 20 and analyzed by ICP-OES (iCAP 7400 Radial; Thermo Scientific) using EPA method 6010B and a PQL of 0.1 mg·kg−1 for As, Cd, and Pb. The ICP-OES equipment was not available during the period of analysis for Expt. 3, and 0.25-g samples were therefore tested by ModernCanna Laboratories using EPA microwave-assisted digestion method 3051A and analysis by ICP-MS (Agilent 7800; Agilent) using EPA method 6020A with a PQL of 0.1, 0.05, and 0.1 mg·kg−1 for As, Cd, and Pb, respectively. Because of extraction and equipment differences, this limits comparison of data from Expt. 3 and from Expts. 1 and 2, although the PQL limits were similar.
In all three experiments, analytical blanks were prepared using coconut coir substrate that was oven dried at 70 °C and finely ground before analysis. The available plant tissue standard reference materials (SRMs) did not contain certified concentrations of all three analytes of interest (As, Cd, and Pb). For this reason, coconut coir laboratory fortified matrices (LFMs) were generated and used to quantify recovery for quality assurance. In Expt. 1, the LFMs were made from reagent grade CdCl2 (Alfa Aesar) and PbCl2 (Acros Organics), and in Expts. 2 and 3, the LFMs were made from National Institute of Standards and Technology standard solutions (lot numbers 1359630, 130116, and 101026, respectively). Applied Cd and Pb concentrations in the LFMs from Expt. 1 were 0.48, 0.97, 4.84, and 9.68 mg·kg−1, and mean recoveries (means ± standard deviation) were 79 ± 22% and 81 ± 18% for Cd and Pb, respectively. The LFMs used in Expt. 2 had applied concentrations of 3.0, 15.0, 30.0, and 60.0 mg of Cd and Pb per kg of dry coir, and mean recoveries were 101 ± 10% and 112 ± 8% for Cd and Pb, respectively. Expt. 3 LFMs included applied concentrations of 0.5, 1.0, 3.0, and 5.0 mg As, Cd, and Pb/kg dry coir, and mean recoveries were 134 ± 22%, 114 ± 54%, and 119% ± 31% for As, Cd, and Pb, respectively. Typical recovery guidelines for method validation are between 80% and 120% (Todorov et al. 2018). The LFM recoveries met these criteria for all analytes in Expts. 1 and 2, and for Cd and Pb in Expt. 3, with greater uncertainty on As measurement in Expt. 3.
Environmental data were collected continuously using data loggers for all three experiments and are summarized in Table 1. Substrate pH and EC were also monitored weekly using a saturated paste extract in Expt. 1 (Warncke 1995) and the press extraction method (Scoggins et al. 2002) in Expts. 2 and 3 (Table 1). Substrate data included bulk density and container volume, for later use in calculating the mass (mg) of HM applied to the root zone. Plant data included dry mass to measure plant growth and calculate the mass (mg) of HM absorbed per plant. Crop leaf greenness was assessed using a soil–plant analysis development (SPAD) chlorophyll meter by taking the average of five measurements per plant from recently mature leaflets (Konica Minolta, Tokyo, Japan).
Analysis in all experiments of As, Cd, or Pb in shoot tissue was by HM. A one-way analysis of variance (ANOVA) was used to assess the impact of HM concentration in the substrate on SPAD, dry mass, Cd and Pb concentration in the plant tissue (mg/kg), and Cd and Pb uptake by the plant (mg/plant) for Expt. 1. A two-way ANOVA with interactions was used to assess the impact of HM concentration and plant species on SPAD; dry mass; As, Cd, and Pb plant tissue concentration (mg·kg); and As, Cd, and Pb uptake in the plant (mg/plant) for Expts. 2 and 3.
Shoot dry mass and SPAD chlorophyll index decreased as HM concentration in the substrate increased (P < 0.001; Fig. 1A and 1C) in Expt. 2, which included high applied concentrations of Cd and Pb. However, there was no effect of HM on these plant parameters in Expts. 1 or 3, in which applied HM concentrations in the substrate were lower. Visible symptoms of heavy metal toxicity were observed in the 30 and 60 mg·kg−1 HM treatments in Expt. 2 (with mustard shown in Fig. 2), but it is unclear whether these symptoms were caused by Cd or Pb, because the two metals were applied together, or an induced micronutrient deficiency. The reductions in plant dry mass and toxicity symptoms observed in Expt. 2 are consistent with previous research on HM uptake in plants. Studies with Salvinia cucullata (a macrophyte species used for HM bioassays) found decreased biomass and chlorophyll content with increased concentration of Cd and Pb, and visible toxicity symptoms were reported in a separate experiment when plants received HM applications with concentrations similar to Expts. 2 and 3 (Phetsombat et al. 2006). Another study (Shi et al. 2012) with multiple hemp accessions resulted in similar biomass reductions and toxicity symptoms from Cd exposure. In Expt. 3, shoot dry mass was not affected by HM applications, but dry mass differed between species (0.34, 0.26, and 0.20 g for mustard, kale, and hemp, respectively).
Citation: HortScience 60, 5; 10.21273/HORTSCI18496-25
Citation: HortScience 60, 5; 10.21273/HORTSCI18496-25
Lead concentrations in shoot tissue were consistently low compared with Cd (all experiments) and As (which was only applied in Expt. 3), indicating that Pb taken up and translocated to shoots was lowest of the three HMs tested. In the three experiments, HM concentrations were only measured in aboveground tissue, because the focus was on leaves, stems, and flowers, which are the commercially harvested portions of hemp. Aboveground HMs are referred to in this study as “plant available,” recognizing that HM concentrations including Cd, Pb, and As can be an order of magnitude higher in roots of Cannabis than in stem and leaf tissue because of limited translocation (Ćaćić et al. 2019; Citterio et al. 2003). In Expts. 1 and 3, Pb tissue concentration (mg·kg−1) and uptake (mg/plant) were unaffected by the HM concentration applied to the substrate. In both of these experiments, Pb concentration in hemp tissue averaged 0.1 mg·kg−1 (Figs. 3A and 5C), which was below the regulatory threshold of 0.5 mg·kg−1. In Expt. 2, with higher applied HM concentrations up to 60 mg·kg−1, Pb uptake (mg/plant) and tissue concentration did increase with increasing HM concentration in the substrate (P < 0.001) and reached a mean tissue Pb concentration of 2.97 mg·kg−1. Hemp accumulated more Pb compared with mustard and kale (Fig. 4D), but average Pb concentrations in hemp tissue remained below the regulatory threshold of 0.5 mg·kg−1 until substrate Pb concentrations reached 30 or 60 mg·kg−1 in Expt. 2.
Citation: HortScience 60, 5; 10.21273/HORTSCI18496-25
Citation: HortScience 60, 5; 10.21273/HORTSCI18496-25
Citation: HortScience 60, 5; 10.21273/HORTSCI18496-25
Only a small proportion of Pb was taken up relative to the mg of HM applied per plant. In Expt. 2, a maximum of 0.3% of the Pb applied to the substrate was absorbed into shoot tissue. The low plant availability of Pb compared with Cd and As is consistent with previous research examining HM accumulation in plants (Rezvani and Zaefarian 2011; Sun et al. 2009). Low Pb availability in the substrate is probably in part a result of root zone pH, with Pb solubility increasing as pH decreases (Karimi 2017; Martínez and Motto 2000). Earlier research suggests that mobile Pb2+ species are mostly stable at pH conditions less than 2.0 in oxidizing environments, further explaining the lack of Pb uptake (Hermann and Neumann-Mahlkau 1985). In addition, the substrate may have reduced HM availability through adsorption onto cation exchange sites (Pilon-Smits 2005).
Arsenic, which was applied only in Expt. 3, increased in plant tissue concentration and uptake per plant with increasing concentrations of applied As to the substrate (P < 0.001). Average As concentrations in hemp tissue were above the regulatory threshold of 0.2 mg·kg−1 for every treatment group except the 0 mg·kg−1 control plants (Fig. 5A), which suggests that As is more plant available than Pb in substrate. Fig. 5D shows that mustard, kale, and hemp shoot tissue contained maximums of 7.8%, 6.1%, and 2.6% of the applied As, respectively, indicating that the two brassica species were more effective accumulators of As compared with hemp. The substrate exchange and pH trends differ for As availability compared with Cd or Pb, because As usually exists in solution as an anion, and As solubility typically increases with rises in substrate pH in oxidizing environments (Beesley and Marmiroli 2011; Fitz and Wenzel 2002; Lee et al. 2015; Moreno-Jiménez et al. 2012; Shen et al. 2020; Smith et al. 1999). In the context of horticultural production, liming agents used to increase substrate pH also play a key role in modulating the solubility of As species, with carbonates, sulfates, and cations such as calcium potentially forming precipitates with As species and removing them from the soil solution where they could be absorbed by plants (García-Sánchez et al. 2010; Lee et al. 2015).
Cd was the most plant-available HM tested in all three experiments (Figs. 3B, 4C, and 5E). Plant uptake (concentration and mg/plant) of Cd increased as Cd concentration in the substrate increased in all three experiments (P < 0.001). Mustard accumulated the most Cd, followed by kale and hemp. A maximum of 32% of the applied Cd was present in mustard shoot tissue, compared with 5.7% for kale and 2.5% for hemp. Plant tissue Cd concentrations reached a maximum of 92.6, 39.9, and 10.4 mg·kg−1 for mustard, kale, and hemp, respectively. In all three experiments, average Cd concentrations in the plant tissue (all treatments other than the control group) were above the regulatory threshold for hemp of 0.2 mg·kg−1. These three experiments are consistent with previous research showing that Cd has greater bioaccumulation than As and Pb in plants (Rezvani and Zaefarian 2011; Sun et al. 2009; Vasile et al. 2021). The increased availability of Cd in the rhizosphere is probably a result of Cd being highly soluble across a wide pH range (Hermann and Neumann-Mahlkau 1985; McBride et al. 1997; Shen et al. 2020).
The conditions in which these three experiments were conducted potentially increased plant availability of As, Cd, and Pb in coconut coir substrate compared with other root zone conditions that may occur in hemp production. Coconut coir has a low cation exchange capacity (CEC) compared with field soils (Landis et al. 2014), resulting in a greater concentration of HM in the soil solution due to the lower ability of coir to tightly adsorb HM onto exchange sites. Application of dissolved HM into the substrate would also be expected to enhance the availability of HM for plant uptake, because it ensured that the majority of HM present in the root zone would initially be in the mobile fraction of the substrate, in contrast to native soils that would be more likely to have HM embedded within organic and mineral particles (Zwolak et al. 2019). Acidic substrate pH ranges were also maintained for the duration of each experiment (Table 1), which probably contributed to greater solubility of HM in the root zone and thus increased plant uptake. Due to the experimental factors aimed at ensuring HM uptake, it is probable that As, Cd, and Pb would be less plant available in soilless substrates in which HM are present but only within the substrate organic and mineral particles, although acidic root zone conditions would increase the risk of HM contamination (Javid et al. 2012). A limitation of this study is that only one soilless substrate type was tested, and other components such as peat, bark, and wood fiber substrates have distinct physical and chemical properties that interact with HM plant availability.
In terms of minimizing HM accumulation, substrate amendments that increase CEC and raise pH are two of the most practical management tools (Sánchez‐Martín and Sánchez‐Camazano 1993; Shen et al. 2020; Zwolak et al. 2019). Amendments that are rich in carbon and organic matter, such as biochar, increase substrate CEC and consequently result in HM being more tightly adsorbed onto exchange sites (Vannini et al. 2021). Heavy metal solubility is also influenced by pH, with many HM species including Pb and Cd decreasing in solubility as pH increases (Karimi 2017; Shen et al. 2020).
Cadmium showed the greatest plant availability, followed by As and then Pb. This trend is consistent with previous research where Cd has shown high mobility across a wide pH range and greater bioaccumulation in plants than other HM (Hermann and Neumann-Mahlkau 1985; McBride et al. 1997; Rezvani and Zaefarian 2011; Sun et al. 2009; Vasile et al. 2021). The low plant availability of Pb is consistent with research on other plant species such as Aeluropus littoralis and Bidens pilosa L. (Rezvani and Zaefarian 2011; Sun et al. 2009) and is notable given some of the high concentrations of Pb found in common soilless substrates (Abreu et al. 2005; Fritz and Venter 1988; Livett et al. 1979; Smieja-Krol et al. 2010). The lack of Pb in shoot tissue, even with experimental conditions that included a low CEC substrate, application of soluble Pb solution, and an acidic substrate pH of 5.6 to 6.3, suggest that Pb had limited root absorption and/or translocation into aboveground tissue. Despite the application of soluble HM to the root zone, only a fraction of the applied HM were absorbed by the plants, which suggests that HM can be removed from the soil solution by the substrate through stabilization processes such as adsorption. Further research could examine the plant availability of these HM in other common greenhouse substrates that vary in CEC, such as peat, bark, and wood fiber, or in hydroponic solution.
Of the three plant species tested, mustard demonstrated the greatest capacity for As and Cd uptake, followed by kale and then hemp. Hemp showed the greatest accumulation of Pb across all three experiments; however, the quantity of Pb taken up into the aboveground tissue was marginal and below most regulatory standards. Plant uptake of HM by mustard, kale, and hemp emphasize that HM contamination is a potential concern for both medicinal and food crop quality.
Bioassay experiments with plant parameters including shoot dry mass in Expt. 2 (A) and Expt. 3 (B) and soil–plant analysis development (SPAD) chlorophyll index for Expt. 2 (C) and Expt. 3 (D). The letters indicate mean comparison for the interaction of species and applied heavy metal concentration using Tukey’s honestly significant difference at α = 0.05.
Visual symptoms of heavy metal toxicity in mustard, which were only observed with the 30- and 60-mg/kg Cd and Pb treatments in high-dose bioassay (Expt. 2).
Greenhouse Expt. 1 with Cd and Pb concentration in shoot tissue (A) and proportion of applied Cd and Pb taken up into shoot tissue per plant [B, heavy metal (HM) shoot concentration × shoot dry mass/HM applied per plant]. In comparison with the measured concentrations in (A), the Florida regulatory thresholds are 0.5 mg/kg for Pb and 0.2 mg/kg for Cd. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05 by heavy metal.
Uptake of heavy metal (HM) into shoot tissue in the high-dose bioassay (Expt. 2). (A and B) Average Cd (A) or Pb (B) tissue concentration by HM treatment level and species. (C and D) Proportion of applied Cd (C) or Pb (D) taken up into shoot tissue (HM shoot concentration × shoot dry mass/HM applied per plant). Note that the scale of the y axis varies between graphs. The dashed lines on the tissue concentration graph (B) indicates the Florida regulatory threshold for Pb of 0.5 mg/kg, and the threshold for Cd in graph (A) is 0.2 mg/kg. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05.
Uptake of heavy metal (HM) into shoot tissue in the low-dose bioassay (Expt. 3). (A to C) Average As (A), Cd (B), and Pb (C) concentrations in shoot tissue by HM treatment level and species. (D to F) The proportion of applied As (D), Cd (E), and Pb (F) taken up into shoot tissue by HM treatment level and species. The dashed lines on the tissue concentration graphs (A) and (C) indicate the Florida regulatory threshold for As and Pb, and the threshold for Cd is 0.2 mg/kg. Note that the scales of the y axes vary between graphs. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05.
Contributor Notes
Funding was provided by industry partners in the Floriculture Research Alliance (floriculturealliance.org) including Quality Analytical Laboratories, the USDA National Institute of Food and Agriculture projects multi-state NC1186 and HATCH FLA-ENH-005918.
Bioassay experiments with plant parameters including shoot dry mass in Expt. 2 (A) and Expt. 3 (B) and soil–plant analysis development (SPAD) chlorophyll index for Expt. 2 (C) and Expt. 3 (D). The letters indicate mean comparison for the interaction of species and applied heavy metal concentration using Tukey’s honestly significant difference at α = 0.05.
Visual symptoms of heavy metal toxicity in mustard, which were only observed with the 30- and 60-mg/kg Cd and Pb treatments in high-dose bioassay (Expt. 2).
Greenhouse Expt. 1 with Cd and Pb concentration in shoot tissue (A) and proportion of applied Cd and Pb taken up into shoot tissue per plant [B, heavy metal (HM) shoot concentration × shoot dry mass/HM applied per plant]. In comparison with the measured concentrations in (A), the Florida regulatory thresholds are 0.5 mg/kg for Pb and 0.2 mg/kg for Cd. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05 by heavy metal.
Uptake of heavy metal (HM) into shoot tissue in the high-dose bioassay (Expt. 2). (A and B) Average Cd (A) or Pb (B) tissue concentration by HM treatment level and species. (C and D) Proportion of applied Cd (C) or Pb (D) taken up into shoot tissue (HM shoot concentration × shoot dry mass/HM applied per plant). Note that the scale of the y axis varies between graphs. The dashed lines on the tissue concentration graph (B) indicates the Florida regulatory threshold for Pb of 0.5 mg/kg, and the threshold for Cd in graph (A) is 0.2 mg/kg. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05.
Uptake of heavy metal (HM) into shoot tissue in the low-dose bioassay (Expt. 3). (A to C) Average As (A), Cd (B), and Pb (C) concentrations in shoot tissue by HM treatment level and species. (D to F) The proportion of applied As (D), Cd (E), and Pb (F) taken up into shoot tissue by HM treatment level and species. The dashed lines on the tissue concentration graphs (A) and (C) indicate the Florida regulatory threshold for As and Pb, and the threshold for Cd is 0.2 mg/kg. Note that the scales of the y axes vary between graphs. The letters indicate mean comparison using Tukey’s honestly significant difference at α = 0.05.
