Sodium alginate was selected as a hydrogel medium because it is biologically non-reactive, non-toxic, sustainably sourced, and inexpensive (Tummino et al., 2020). Glomalin, a mixture of proteins and polysaccharides referred to as GRSFs (glomalin-related soil fractions), was selected as a potential stabilizing agent. It is exuded by arbuscular mycorrhizal fungi and is associated with soil aggregate formation enhancing the stability of soil structure (Rillig, 2004). In-vitro experiments were performed to test the impact of added glomalin on the cohesivity of the generated immobilization beads (termed biobeads), the total remediation of orthophosphate (1.5 ppm PO₄ ^3-) and total phosphate (4 ppm PO₄ ^3-), and crop health at a BSL-1 laboratory (Canterbury School, Fort Myers, FL). Biobeads were constructed from a colloidal gel suspension consisting of a 2% sodium alginate solution complex, different concentrations of glomalin, and a culture of C. vulgaris.

Glomalin was extracted from unamended garden soil covered with a thick layer of accumulated leaf litter. Such an environment is conducive for high glomalin production. A soil solution of 1.2 g soil in 10 ml of 50 mM sodium citrate was prepared following the methods of Wright and Upadhyaya (1996) to extract glomalin from the soil. The solution was autoclaved for 75 min at 121°C then centrifuged at 3500 RPM for 15 min. The supernatant was extracted, autoclaved, and centrifuged eight times until the solution changed from opaque to transparent brown. Ten samples were simultaneously processed at each iteration to generate 50 ml of glomalin solution. The C. vulgaris live culture (Suncoast Marine Aquaculture) was agitated to homogenize the C. vulgaris sample before extraction. A diluted solution of food-grade sodium alginate (The Seaweed Solution) was prepared by adding 5 ml of alginate to 100 ml distilled water. The solution was stirred slowly for 30 min, to avoid gas bubbles that destabilize the final structure. The solution was stored for 24 h at 4°C. Secondary solutions were made for the inoculated beads in 20 ml batches. Six groups were determined based on a range of proportional concentrations of sodium alginate, glomalin, and Chlorella vulgaris: 100A-0G-0C, 50A-0G-50C, 50A-10G-40C, 50A-25G-25C, 50A-40G-10C 0A-0G-0C (Table 1).

Approximately 0.2 ml of the secondary solution was added dropwise to 3% CaCl₂ solution kept on ice to induce spherization. The generated beads averaged between 4 and 6 mm. Beads were left in the CaCl₂ solution for 30 min to mature, drained from the CaCl₂ solution using a nylon mesh, rinsed with distilled water, counted, and placed in the nutrient solution (Gonzalez and Bashan, 2000). During the spherization process, bead formation results from both intra-molecular interactions (between the polysaccharide and organic acid matrix) and the sodium/cation exchange generated in the solution. The resulting bead is temporarily insoluble yet permeable in water and can be inoculated with a variety of microorganisms or enzymes that maintain their biological functionality within the medium (Aydelotte et al., 1998; Jin et al., 2011). Although the alginate beads are water insoluble, they are pH and cation sensitive and can lose their cohesivity over time (Velings and Mestdagh, 1995). The impact of the glomalin/alginate matrix on the vitality of C. vulgaris was assessed in eutrophic concentrations (∼1.5 ppm PO₄ ^3-) and hypereutrophic concentrations (∼4 ppm PO₄ ^3-) to determine if excessive eutrophication would destabilize the biobeads or inhibit the functionality of C. vulgaris in the remediation of phosphate. The bioavailable form of P, orthophosphate was measured under both conditions to determine the viability of the system. Total P was measured in the second set of experiments to ensure that the drop in orthophosphate concentration was associated with metabolic uptake, not P transformations.

A P solution (Carolina Biological) was diluted with distilled water at 10 ml L⁻¹. Twenty-five biobeads were added into 200 ml of solution per flask: 100A-0G-0C, 50A-0G-50C, 50A-10G-40C, 50A-25G-25C, 50A-40G-10C 0A-0G-0C. A low range phosphate colorimeter (Hanna Instruments HI 713–0; 0.01 ppm resolution and+/−0.03 ppm accuracy) was used to determine the concentration of orthophosphate in the nutrient solution. The colorimeter was used in conjunction with pre-packaged reagents in accordance with the 4500-PE ascorbic acid method which can determine a range in orthophosphate concentrations from 0.1 to 6 ppm. 10 ml water samples were collected every 3 days after start of the experiment up to 9 days (3 DAT, 6 DAT, and 9 DAT).

A hypereutrophic P solution was used to determine whether an increase in P destabilize the biobeads or if the system would be inhibited by P toxicity. A 1000 ppm TP solution (Fischer Scientific) was diluted to ∼4 ppm. 400 ml of nutrient solution was added to thirty 500 ml beakers designated as: 100A-0G-0C, 50A-0G-50C, 50A-10G-40C, 50A-25G-25C, 50A-40G-10C, and 0A-0G-0C. Fifty beads were added to each beaker and two 10 ml samples were extracted every 7 days for 5 weeks. One 10 ml sample was analyzed immediately for orthophosphate using the heteropoly molybdenum blue method (Holman, 1943) with a high range phosphate colorimeter (Hannah Instruments HI 717; 0.1 ppm resolution and accuracy of+/-1 ppm). The second 10 ml sample was stored at 4°C for analysis. Total P was analyzed using an inductively coupled plasma-optical emission spectrometer (5110 ICP-OES, Agilent Technologies Inc, CA) (U.S. EPA. 1994) at the University of Florida Soil, Water, and Nutrient Management Laboratory in Belle Glade, FL.

Throughout both experiments, the internal cohesivity of the beads was assessed by determining the pressure the bead could withstand before complete structural collapse using a penetrometer. Every 7 days, three beads were removed from the system and placed within the wells of a microwell plate. The pressure necessary to structurally destabilize the biobeads was measured by a penetrometer in kg cm⁻² (105 Pa). A penetrometer (sclerometer, model GY-1), with a range of 2–15 kg cm⁻² (x 105 Pa) and a resolution of+/-0.01, was used to determine the force necessary to crush the beads in kg cm⁻² (105 Pa).

Germination, plant growth, leaf number, and dry biomass were assessed to determine if the application of the beads impacted crop growth. The P loaded biobeads were applied to the surface of sterile, hydrated soil pods approximately 4 cm in diameter and 6 cm high (Barefoot Farmer organic farm, TN) and planted with Brassica rapa (subspecies Kosaitai raab) seeds. A moisture wick was run through the pods which were then rehydrated using DI water. Three B. rapa seeds were placed in each soil pod. Five beads from each concentration group were placed in 10 pods and 10 pods received no beads as a control. The soil was mixed with a scoopula to 1 cm depth with the beads intact. The pods were separated into five hydration trays and distributed in the trays in a randomized block design. The hydration trays were placed in a plant growth cart equipped with 2 grand lux tubes and 2 incandescent grow bulbs (60 W) adjusted for a 10 h photoperiod (Carolina Biological, NC). Seed germination was assessed after 10 days. Growth from the soil surface to the base of the longest terminal leaf was measured at 10 and 34 days. At 34 days, the stems were cut at the surface of the soil and the number of leaves was counted per plant. The plants were grouped according to treatment, placed on a drying surface, and dried in an oven for 4 days at 40°C. Dry biomass was measured for each group.

The effects of added glomalin and incubation time on the cohesivity of biobeads, the total remediation of orthophosphate and total phosphate, and crop health was analyzed using randomized complete design analysis with the generalized linear model (GLIMMIX) procedure in SAS (Version 9.4; SAS Institute Inc. Cary, NC). The treatment × date effect was significant; thus, data were analyzed separately for each date. Differences between treatments at each incubation time were further evaluated for significance using the Fisher LSD post hoc test. All tests were considered significant at p < 0.05.

An initial experiment to determine if the glomalin alginate matrix would inhibit the bioremedial abilities of C. vulgaris was conducted using a phosphate solution of 1.68 ppm. Both time and treatment had a highly significant effect on the reduction of orthophosphate from the solution (Table 2). The positive control group (alginate-only beads) significantly lowered orthophosphate concentrations compared to the negative control group (glass beads) at 6 DAT and 9 DAT (Table 3). However, there was no difference between two control groups at 3 DAT. At 3 DAT, each glomalin/Chlorella bead had lower mean concentrations of orthophosphate compared to the alginate-only beads and glass beads in each measured date except group containing 50% sodium alginate, 0% glomalin, 50% C. vulgaris. All alginate bead types reduced orthophosphate concentration over time. At 9 DAT, alginate-only beads extracted 43% of the orthophosphate from the system. Alginate beads with glomalin and Chlorella reduced the orthophosphate concentration more efficiently than alginate-only beads and no-alginate beads. The most significant drop in orthophosphate was observed in the 50A-10G-40C groups which had an average concentration reduction from 1.68 to 0.17 ppm, or 90%, over the 9 days of observation. Significant lower orthophosphate concentration in positive control group compared to negative control group at 6 DAT and 9 DAT suggests that the alginate bead itself absorbs the solution and sequesters orthophosphate even in the absence of a bioremedial organisms. A lower mean concentration of orthophosphate in the solution between each of the glomalin/C. vulgaris beads compared to the alginate-only beads and no-alginate beads indicates that the beads with C. vulgaris had additional effects in the remediation of orthophosphate. Lau et al. (2010) also reported alginate immobilized C. vulgaris had higher nutrient removal efficiency from wastewater than free cells.

The change in orthophosphate concentration was analyzed over time to determine if the removal of orthophosphate was constant or if an assimilation lag would be observed. No significant change in orthophosphate concentration was observed for negative control over time. No significant change in orthophosphate concentration was observed for all alginate beads (Table 4) up to 9 DAT. However, there was a significant reduction in orthophosphate concentration at 30 and 35 DAT for all alginate beads groups compared to the initial concentration. The most significant drop in orthophosphate was observed in the groups containing 50% sodium alginate, 25% glomalin, 25% C. vulgaris which had an average concentration reduction of 51% over the 35 days of observation. At 0 DAT, a significant difference in orthophosphate concentration was observed between treatments. At 9, 16, 23, 30, and 35 DAT, the negative control group had higher orthophosphate concentrations than positive control group and rest of the alginate group with glomalin/C. vulgaris except positive control, group containing 50A-10G-40C, and group containing 50A-40G-10C at 23 DAT (Table 4). No significant difference was measured in the mean reduction of orthophosphate between most of the biologically active groups, reinforcing the finding that the addition of glomalin did not inhibit the bioremedial ability of C. vulgaris. There was no significant difference between the mean orthophosphate concentrations up to 9 DAT, but there was a significant reduction at 30 and 35 DAT, indicating higher levels of eutrophication may increase the time needed for microorganisms to digest the P, thus causing a lag in nutrient remediation. Rasoul-Amini et al. (2014) investigated the influence of three microalgae strains (Chlamydomonas sp, Chlorella sp, and Oocystis sp.) on N and phosphate removal and found all microalgal strains achieved their best removal efficacies for PO₄ ^3- at 14 DAT, though that was also the last day of testing for the study. Therefore, a longer culture time appears to increase microalgae nutrient absorption. The lack of significant differences in the mean reduction of orthophosphate between most of the biologically active groups indicates the addition of glomalin did not inhibit the bioremedial ability of C. vulgaris.

A significant reduction in the TP concentrations was measured between sampling dates (0 DAT to 35 DAT) except for negative control (Table 5). There were no significant differences in TP concentration between the glomalin/Chlorella beads and the positive control except group containing 50A-25G-25C at 16 DAT, group containing 50A-25G-25C at 23 DAT, and group containing 50A-0G-50C at 30 DAT. Although the binding of phosphate in the positive control was observed, the drop in TP in the solution was not statistically significant between the positive and negative control except at 9 DAT and 30 DAT. The glomalin/C. vulgaris groups, regardless of glomalin concentration, significantly reduced the TP in the solution compared to the negative control. AT 30 DAT, the negative control had significantly higher TP compared to other alginate groups. No significant differences in TP concentration between the glomalin/C. vulgaris groups and the positive control confirms that the sodium alginate matrix is capable of absorbing phosphate without the introduction of bio-remedial organisms. Mohseni et al. (2021) reported alginate beads of C. vulgaris were sufficiently stable and were capable of remediation of TP up to 100% over a 10-day period. The lack of significant variance between the means of the biologically active groups indicates that the introduction of glomalin did not inhibit the ability of C. vulgaris to remove P from the solution. In addition, although a greater loss of TP was observed in biologically active groups, this difference was not statistically significant.

All glomalin/C. vulgaris groups displayed increased structural cohesivity over the experimental period. In all groups, there was a significant increase in the bead resilience to pressure over time (Table 6). A significant increase in cohesivity was observed over the first 14 days but not between days 14–28 for positive control group and groups containing 50% sodium alginate, 40% glomalin, 10% C. vulgaris whereas groups containing 50% sodium alginate, 10% glomalin, 40% C. vulgaris and 50% sodium alginate, 25% glomalin, 25% C. vulgaris no significant difference in cohesivity was observed over the first 14 days (Table 6). The group with the greatest relative concentration of glomalin exhibited a statistically significant increase (208%) in cohesive stability. However, while relationship between glomalin concentration and structural integrity was linear, the relationship between glomalin concentration and biobead strength was parabolic. This parabolic relationship between glomalin concentration and biobead strength suggests that an increase in glomalin may initially inhibit the cross-linkages that increase gelation and cohesivity. However, at higher concentrations interaction with the glomalin may increase intra-bead molecular interactions, increasing structural strength. The lack of significant differences in cohesivity over the first 14 days suggests there is a maturation period of 2 weeks during which the alginate beads continue to experience gelation processes.

Seed germination was assessed by counting the number of sprouted seeds at day 10. Although on average there was a greater percentage of seed germination between the groups containing glomalin and Chlorella, there was no significant difference in germination between treatments (Figure 1). There was no significant difference in growth at 34 days or biomass between the treatment groups (Figure 2, Figure 3). The leaf number was significantly higher in the high concentration glomalin groups (40% glomalin and 25% glomalin) than in the low concentration group (10% glomalin) and controls. The lack of significant differences between treatment groups for the seed germination, biomass, and plant growth suggests the application of the biobeads did not confer a growth advantage to the B. rapa seeds. However, the higher leaf numbers in the high concentration glomalin groups (40% glomalin and 25% glomalin) compared to the low concentration group (10% glomalin) and the control groups suggests increased concentrations of glomalin related soil fractions (GRSFs) may improve leaf growth.