MARZ is a soil remediation experiment within Metabolic Studio’s Un-development 2, a project that aims to transform an industrial warehouse into a living riverbed connected to the Los Angeles River floodplain. The site was formerly associated with Standard Oil and then used as a petrochemical storage facility, and baseline testing confirmed elevated lead and petroleum-range organics beneath the warehouse floor.
From April 2023, to June 2026, artist/chemist Maru García worked with Metabolic Studio’s Farmlab team, developing protocols and testing bioremediation strategies that could be used in situ. The first phase of the project tested surface-applied zeolites, compost, mulch, and phytoremediation, with three different treatment protocols in nine plots. In the second phase, after identifying willow as a strong remediator, we favored willow phytoremediation and added oyster mushrooms in all plots.
The study prioritized methods that reduce exposure, avoid disruptive excavation, support soil as a living ecosystem, and can be replicated at sites without industrial remediation infrastructure. We have included our results and methodology below along with a poster summarizing the results, and a remediation toolkit that explains how you can use these methods yourself.
Click here to download the poster and the soil remediation toolkit.
Phase 1: Lead Remediation Using Natural Treatments
WHAT IS IT?
The warehouse containing Un-development 2 was built in the 1970s, on a site that was once home to a Standard Oil facility. Before industrialization, however, the land was part of the unbridled Los Angeles River floodplain on the ancestral lands of the Tongva, Gabrieleno Shoshone, and Kizh nations. Un-development 2 reclaims that thriving riverbed, harmonizing the site’s geological past and speculative future by uncovering and restoring the compromised soil beneath the concrete.
Part of our Un-development 2 project, which spans three warehouse spaces, involved designing and executing experiments to remove petrochemical contamination, lead and arsenic in particular, stored in the soil underneath the warehouse floors. The project objectives and considerations for this first phase were:
Study the application of natural materials for lead remediation in contaminated soil
Use a method safe for humans and the environment
Use a method that is easily replicable for other contaminated sites where industrial processes are unavailable (e.g., backyards and parks)
Minimize soil disturbance to reduce exposure to contaminants
Value soil as an ecosystem that needs care and healing
HOW DID WE DO IT?
Site Preparation
For Phase 1, we removed the concrete from a section of the middle warehouse, now called MARZ. The study site was closed to the public and was divided into nine plots to test each treatment in triplicate. Three samples were taken from each plot and tested as a composite sample. The results above, sorted by plot, showed some lead levels above the toxic waste limits for total lead, soluble lead, and diesel and oil organics.
Remediation Methods Tested
Over a period of 18 months, we tested the “least disturbance soil remediation” (LDSR) method and phytoremediation. LDSR is a DIY method used to reduce lead exposure in residential sites, where the priority is to reduce dust production and minimize exposure to contaminants for those involved in remediation efforts. LDSR applies layers of amendments to the surface of the affected area, including zeolites, compost, and mulch.
Zeolites are a mineral with a unique molecular structure. When zeolites come into contact with water, the pores of the mineral become hydrated. This creates the conditions for an ion exchange, and the sodium or calcium naturally present in the zeolite is replaced by lead. This encapsulates the lead; it is no longer bio-available and therefore won’t be absorbed into the bloodstream. Constant watering also moves the amendments through the soil layers. Compost and mulch attract and carry microorganisms capable of breaking down petroleum contaminants, increasing soil health, and providing a physical barrier.
Treatments applied to MARZ, the isolated section of the industrial warehouse, are as follows:
Treatment A: zeolites + compost + mulch + corn + mycorrhizal fungi + willow + watercress
Treatment B: zeolites + compost + mulch
Treatment C: compost + mulch
Treatment 3C: compost + mulch + floodplain
We selected plants that would thrive in the conditions inside the warehouse: low light and high humidity, and have a demonstrated ability to bioaccumulate lead including arroyo willow, California mugwort and watercress. We also discovered volunteer plants: milk thistle, morning glory, horseweed, and sow thistle. They likely entered the site as seeds in the compost or mulch amendments, and were able to thrive in the low light and high humidity.
RESULTS
Lead
The results for all treatments in the six sampling sessions are presented above. We observed an average total lead reduction, by 41% in plots receiving Treatment A, and by 27% in plots receiving Treatment B.
In the plots receiving Treatment A, we expected to find that the introduced plants bioaccumulated lead in their roots, resulting in a total lead reduction over time. Although we only incorporated plants into the remediation process for plots receiving Treatment A, throughout the experiment, plants naturally grew in other plots as well. This could help explain the average total lead reduction of 27% in plots receiving Treatment B, which was a more significant reduction than we were expecting for zeolites, compost and mulch alone.
Although we measured the average reductions in plots receiving both treatment A and B, we noted that each plot exhibited high variability in lead concentration, leaving hot spots.
We separately monitored the lead concentration in plant roots and in plant tissue above ground (leaves and stems). As demonstrated above, we found that lead is primarily stored in the root systems, and arroyo willow and sow thistle were particularly effective bioaccumulators. The plants’ above-ground tissue contained lead in a much lower concentration, with lead extraction levels below 10 ppm.
We also analyzed levels of soluble lead, also known as bioavailable lead. This measurement allows us to determine whether the lead in the soil will be easily absorbed by living organisms. While applying zeolites doesn’t reduce the total amount of lead, it does reduce the concentration of soluble lead. Treatment C (compost + mulch) showed the most significant trend in reducing bioavailable lead, followed by Treatment B (zeolites + compost + mulch). They presented an average reduction of 37% and 4%, respectively.
Comparing these results with others published using the same methods, we hypothesized that the compost used in Treatment B and C may be acting as a binding agent, trapping the lead and increasing soil pH, processes that reduce lead bioavailability. It is important to mention that in comparison to zeolites, which permanently absorb lead, organic matter in the compost will degrade and the lead will eventually be released back into the soil.
Arsenic and Cadmium
In the case of other contaminants, we found that arsenic was reduced by 49% in all treatments while cadmium, the results did not show a clear pattern.
Petroleum-derived Contaminants
For petroleum-derived contaminants, we did not apply a direct remediation method besides the microorganisms present in the compost, and we found that the Diesel Range Organics were above the limit in almost all the samplings. We did not find Gasoline Range Organics (C6-C10) in any sample.
WHAT DID WE LEARN?
Our results show a trend in the reduction of total lead concentration in the plots where plants were present, and that applying zeolites or compost could reduce bioavailable lead concentrations in the long term.
In designing our experiments, we mixed approaches, combining plants, amendments and zeolites to allow zeolites to penetrate the ground as the plants’ roots worked their way through the soil. We eventually determined that treatments should only contain one remediation method, to better understand the action of each method.
Treatment A should not contain zeolites to allow the plants to take up the lead. In the case of Treatment B, the zeolites should also be tested without the interference of compost to isolate the results.
Despite both bioremediation and zeolite application successfully removing lead from the soil in MARZ, these treatments are safe but slow-acting. Due to time constraints and other site-based difficulties—heavily-compacted soil and uneven lead distribution for example—we will need to incorporate other solutions to remediate the remainder of the warehouse. The LDSR (least disturbance soil remediation) methods we tested can, however, be very beneficial in other conditions, like reducing lead exposure in residential sites with lower lead levels, where residents will remain in place during remediation. In industrial situations with high lead levels, it would be beneficial to use a more direct approach.
Additional graphed results are available on our website. Do you have any questions about bioremediation or Lease Disturbance Soil Remediation? Let us know in the comments!
Love this experiment.