This survey analyzed feed ingredients, mineral products, premixes and animal feeds collected from Texas manufacturing and distribution establishments between October 2010 and May 2015. The authors report 564 samples, while their fiscal-year table sums to 565. The study measures upstream feed materials, not metals in meat, milk, eggs or dog and cat foods.
Key numbers
| Feed matrix and element | N analyzed | Mean ± SD | Median |
|---|---|---|---|
| Fish meal, total arsenic | 38 | 3.21 ± 1.34 mg/kg | 3.6 mg/kg |
| Fish meal, total mercury | 38 | 0.056 ± 0.029 mg/kg | 0.062 mg/kg |
| Beef cattle feed, chromium | 160 | 4.91 ± 3.92 mg/kg | 4.0 mg/kg |
| Mineral products, nickel | 55 | 56.91 ± 138.38 mg/kg | 15.0 mg/kg |
| Premixes, selenium | 28 | 74.6 ± 69.5 mg/kg | 59.5 mg/kg |
| Mineral products, thallium | 124 | 0.27 ± 1.79 mg/kg | 0.0 mg/kg |
These Table 3 statistics cover fiscal years 2012–2015 and include the authors’ substitution of zero for results below quantitation. A zero median therefore does not mean elemental absence. Element-specific N varies substantially. Feed categories include supplements and concentrates as well as complete rations.
Fiscal-year 2011 mineral products had mean arsenic 1.2 mg/kg, cadmium 1.5 mg/kg and lead 20.5 mg/kg (N=58, Table 2). The source’s 2011 cadmium narrative uses conflicting units and extrema; the table values remain distinct source reports.
Methods (brief)
Investigators used official sampling and chain-of-custody procedures. Feeds were ground through a 0.75 mm screen and split before overnight acid predigestion and microwave digestion. ICP optical emission measured the 2011 samples; ICP mass spectrometry measured 2012–2015 samples. The later method’s quantitation limits were 1 mg/kg for most analytes, 0.01 mg/kg for mercury and 10 mg/kg for copper. A numeric limit for the earlier method was not supplied.
The metal panel and range of sampled feed types expanded during the survey. All below-quantitation results were treated as zero for descriptive statistics. Moisture correction is not specified, and liquid feeds must remain separate from dry-feed matrices. Arsenic, mercury and chromium were not speciated. Pearson correlations describe co-occurrence within feed samples, not transfer to animal tissues.
Evidence fitness
Supports qualified feed-type and fiscal-year occurrence summaries, with assay changes and censoring preserved. Mineral and premix concentrations cannot characterize complete rations without their inclusion rates. The paper identifies added copper sources in some feeds, but it does not establish attributable contaminant fractions or measured human dietary exposure.
Limitations
The paper has several internal inconsistencies. Table 3 gives horse-feed lead median 15 mg/kg alongside mean 0.07 mg/kg and SD 0.28 mg/kg. A molybdenum footnote reports nondetection in groups with positive tabulated summaries. Fish-meal mercury narrative extrema conflict with its histogram, and the premix figure caption misidentifies two analytes. These observations remain flagged rather than silently corrected.
The 2012 detection percentage also conflicts with that year’s tabulated counts. Increasing detection over time accompanies more sensitive instruments, additional analytes and changing feed categories. It cannot by itself establish rising contamination. Nutritional guidance and legal limits quoted in the discussion are secondary citations; they do not establish independently verified compliance or current legal maxima. No paired animal-tissue measurements or clinical outcomes were collected.
Related evidence
- Animal feed and the food chain
- Arsenic
- Cadmium
- Chromium
- Copper
- Lead
- Mercury
- Molybdenum
- Nickel
- Thallium
Update history
No substantive edit history is available in this build. The full commit record is available in git.