Heavy metals in our inputs: the test results we have, and the ones we do not

We hold heavy metal results from Hill Laboratories for four things we sell: Living Soil, worm castings, andesite rock dust and pumice. All four were under the reference values printed on the certificates. Mussel meal, fish meal, kelp, BioPhos and the soft rock phosphates have no Seacliff heavy metal test yet.

Our earlier post, Natural does not mean clean, said only a lab sheet settles what is in a bag. Here are ours. Each certificate is on the lab tests page.

How to read these numbers

All metal figures are mg/kg on a dry weight basis. The lab digests the sample in acid and reports 'totals' in quotation marks, because the method slightly under-estimates the true total for some elements (Hill Laboratories 2023b).

• One sample, one date. A result describes the sample we sent, not every bag before or since. We took the samples; the lab analysed them as received.
• The reference values are printed by the lab. Beside each result the certificate prints values from NZS 4454:2005 (Table 3.1, the New Zealand compost standard) and the BioGro Standards 2009 compost limits (Appendix A, Table A3). The lab notes that a result applies to the sample submitted and does not on its own mean the product meets the standard.
• What we could check. BioGro's current Appendix A carries the same compost limits for these eight metals (BioGro 2026). NZS 4454:2005 is a paid standard we could not open, so its values are quoted as printed.
• Accreditation. On the 2022 Living Soil certificate the heavy metal totals are marked as outside the lab's IANZ accreditation (Hill Laboratories 2022). The 2023 and 2024 certificates carry no accreditation statement.

[DAN SIGN-OFF NEEDED: the certificates name the samples "Living Soil", "Andersite" and "Pumice". None says the sample came from stock that was sold. Confirm each one, and that "Andersite" is the Ultra Paramagnetic Andesite Rock. Dan has confirmed the worm castings sample.]

The results

Reports: 2965941 (Hill Laboratories 2022), 3473401 (Hill Laboratories 2024) and 3204677 for the two March 2023 samples (Hill Laboratories 2023b). A "less than" result was below what the test can detect. We hold other certificates that are not quoted here, because they cover trial batches, samples we cannot yet match to a product, or materials we no longer sell.

Product by product

Living Soil

One sample, reported 4 May 2022. All eight metals were under both sets of printed values. The closest was copper, at 44 mg/kg against the BioGro value of less than 60. The certificate states no recipe version. Read it as the Living Soil of April 2022, not the batch on the shelf today.

Worm castings

One sample of the castings we sell now, reported 28 February 2024. All eight metals were under both sets of printed values, with copper at 24 mg/kg against the BioGro value of less than 60. An earlier sample, from a previous supply that we no longer sell, appears on the 2023 certificate beside the andesite and pumice; it is not shown here. One certificate cannot stand for every batch.

Andesite rock dust and pumice

These went to the lab together in March 2023 (the certificate spells the first "Andersite"). Both were booked in under the lab's compost test, so the values printed beside them are compost values. Those were not written for rock, and passing them proves little. The numbers are low: cadmium was below the 0.02 mg/kg detection limit in the andesite and 0.03 in the pumice. Sneath (2020), citing the International Cadmium Association, gives 0.02 to 0.2 ppm of cadmium for igneous and metamorphic rocks. Both samples sit at or below the bottom of that range. It was a heavy metal test only.

Worked example: what a rock dust dressing adds to a bed

Concentration is not dose. Take the andesite at 2.5 kg per square metre, the crop rate on our listing, worked into the top 150 mm of soil.

• Copper applied: 2.5 kg × 25 mg/kg = 62.5 mg per square metre. The certificate is on a dry weight basis and we treat the rock dust as dry.
• Soil in the top 150 mm: 1 m² × 0.15 m = 0.15 m³. At an assumed dry bulk density of 1.0 t/m³ that is 150 kg.
• Rise in soil copper: 62.5 mg ÷ 150 kg = 0.42 mg/kg per dressing.
• Cadmium was below the detection limit of 0.02 mg/kg, so the most a dressing could add is 2.5 × 0.02 = 0.05 mg, or 0.05 ÷ 150 = 0.0003 mg/kg.

BioGro's limit for copper in soil is 60 mg/kg dry weight (BioGro 2026, Table A3). One dressing adds about 0.42 ÷ 60 = 0.7 % of that. Copper is an essential trace element as well as a metal with a limit (BioGro 2026), so a small addition is not a problem in itself. It does not break down, though. With no losses, ten such dressings add about 4 mg/kg.

What we have not tested

• Mussel Meal and Fish Meal. No Seacliff test. The supplier's nutrient analyses we hold, dated December 2017 and September 2018, did not cover arsenic, cadmium, lead or mercury.
• Acadian Kelp Meal and Tri-Kelp. No Seacliff test.
• BioPhos, Colloidal Soft Rock Phosphate and MSR. No Seacliff test. These matter most. Cadmium is a natural impurity in most phosphate sources, and the New Zealand fertiliser industry holds itself to a voluntary limit of 280 mg of cadmium per kg of phosphorus (Rys 2020). We cannot say where BioPhos or the soft rock phosphates sit against that until cadmium and phosphorus are measured on one sample.
• The blended fertilisers and most other single inputs. Nothing published. The full list is on the lab tests page.

We are asking the suppliers for current certificates and getting a quote from the lab so we can test these ourselves. [DAN SIGN-OFF NEEDED: confirm the supplier requests have been sent before this is published.] The results will go on the lab tests page whichever way they fall.

Supplier documents are not batch tests

We hold three supplier papers with heavy metal lines. Seacliff ordered none of them.

• Acadian Kelp Meal. The maker's typical analysis sheet, dated 2002, gives cadmium, lead, mercury and nickel as less than 1 ppm each and chromium as 1 to 2 ppm (Acadian Seaplants 2002). It describes what the maker expects, not a batch, it is 24 years old, and it gives no arsenic figure.
• Fish Bone Meal. The supplier's certificate of analysis reports mercury at 0.07 ppm against a specification of less than 1 ppm (United Fisheries no date). It has no date or batch number, and no arsenic, cadmium or lead result.
• Neem Seed Meal. The maker's certificate for batch 105083, November 2021, lists arsenic, cadmium, lead, mercury and nickel as "NIL" with no detection limits (Sun Bionaturals 2021). The same sheet reports copper at 14 ppm and zinc at 55 ppm, then lists both as NIL under heavy metals. NIL without a detection limit is not a measurement.

We also hold a supplier's 2021 heavy metal report for oyster shell grit. It stays out of this post until the supplier agrees to it being quoted.

What the rules ask of a seller

Since 7 March 2025 anyone who makes, imports or sells a fertiliser or soil conditioner must identify the chemical contaminants that may be present, heavy metals included, manage them, and keep that under review (MPI 2024, clause 2.1). The notice sets no heavy metal limits and does not require publication. Publishing is our choice. Testing more of the range is the work still to do.

If you grow medical cannabis under licence

The minimum quality standard for medical cannabis sets heavy metal limits of 3.0 ppm arsenic, 0.5 ppm cadmium, 5.0 ppm lead and 0.5 ppm mercury (New Zealand Government 2019, regulation 7). Those limits apply to the ingredient or product tested, not to the growing medium, so do not read the table above against them.

A clean input certificate cannot promise a pass. For cadmium there is no simple direct relationship between the concentration in the soil and what the plant takes up; species, variety and soil factors all shift it (Sneath 2020). Your own product test is the one that counts.

What to do

1. Read the certificate, not the summary. Check the sample name, the date and which figures are "less than".
2. Ask every supplier, us included, for a batch test with detection limits. Treat a typical analysis, a specification or "NIL" as no data.
3. For our untested inputs (mussel meal, fish meal, kelp, BioPhos and the soft rock phosphates), wait for the result or have a sample tested yourself. You may not need the product in the meantime.
4. Get a metals baseline for your own soil. Where to test your soil lists the labs.
5. If you are certified organic, ask your certifier before using any input without a current result.

References

• Acadian Seaplants Limited (2002) Ascophyllum nodosum Kelp Meal and Flour: typical analysis, sheet As.Li.2.1.E.0502. Dartmouth, Nova Scotia: Acadian Seaplants Limited. Supplier document held by Seacliff Organics, listed at /pages/lab-tests; copy on request.
• BioGro New Zealand (2026) BioGro Organic Standards, Appendix A: Residue Levels in Certified Products, Water, Soil and Composts, reviewed 2026. Available at: https://www.biogro.co.nz/s/Appendix_A-djgb.pdf (accessed 4 October 2026).
• Hill Laboratories (2022) Certificate of Analysis 2965941-CPv1: Living Soil, reported 4 May 2022. Hamilton: R J Hill Laboratories Limited. Full certificate at /pages/lab-tests.
• Hill Laboratories (2023b) Certificate of Analysis 3204677-CPv1: Andersite, Pumice, Worm Castings #1, reported 29 March 2023. Hamilton: R J Hill Laboratories Limited. Full certificate at /pages/lab-tests.
• Hill Laboratories (2024) Certificate of Analysis 3473401-CPv1: Worm Castings Sample 08/02, reported 28 February 2024. Hamilton: R J Hill Laboratories Limited. Full certificate at /pages/lab-tests.
• Ministry for Primary Industries (MPI) (2024) ACVM Notice: Fertilisers, Plant Biostimulants and Soil Conditioners. Requirements for compliance with conditions of exemption. Issued 7 March 2024, in force 7 March 2025. Wellington: MPI. Available at: https://www.mpi.govt.nz/dmsdocument/61288-ACVM-Notice-Fertilisers-Plant-Biostimulants-and-Soil-Conditioners (accessed 4 October 2026).
• New Zealand Government (2019) Misuse of Drugs (Medicinal Cannabis) Regulations 2019 (LI 2019/321), regulation 7. Available at: https://www.legislation.govt.nz/regulation/public/2019/0321/latest/whole.html (accessed 4 October 2026).
• Rys, G., on behalf of the Cadmium Management Group (2020) 'Cadmium and New Zealand agriculture and horticulture: a refreshed strategy for long-term risk management', in Christensen, C.L., Horne, D.J. and Singh, R. (eds) Nutrient Management in Farmed Landscapes. Occasional Report No. 33. Palmerston North: Farmed Landscapes Research Centre, Massey University. Available at: https://www.massey.ac.nz/~flrc/workshops/20/Manuscripts/Paper_Rys_2020.pdf (accessed 4 October 2026).
• Sneath, G., on behalf of the Cadmium Management Group (2020) 'Revision of the Tiered Fertiliser Management System for soil cadmium', in Christensen, C.L., Horne, D.J. and Singh, R. (eds) Nutrient Management in Farmed Landscapes. Occasional Report No. 33. Palmerston North: Farmed Landscapes Research Centre, Massey University. Available at: https://www.massey.ac.nz/~flrc/workshops/20/Manuscripts/Paper_Sneath_2020.pdf (accessed 4 October 2026).
• Sun Bionaturals India Private Limited (2021) Certificate of Analysis: Terra Neem Cake Powder, batch 105083, November 2021. Supplier document held by Seacliff Organics, listed at /pages/lab-tests; copy on request.
• United Fisheries Limited (no date) Certificate of Analysis: Calcium Fish Bone MCHC Powder (Animal Grade). Christchurch: United Fisheries Limited. Supplier document held by Seacliff Organics, listed at /pages/lab-tests; copy on request.