Our recommendations are based upon the toxicity of inorganic AsIII, specifically the arsenite ion. Routine water quality analysis available to livestock producers does not distinguish between As species, and, although the less toxic pentavalent forms of As are more likely to occur in surface waters, trivalent As is seen frequently enough in specialized surveys to justify the assumption.86,87 It is suggested that ruminant animals are less susceptible to As than monogastrics.25,88 With the exception of laboratory rodents, however, we were not able to confirm this to be the case, thus we have assumed horses are equally sensitive to As as ruminants.
Chronic poisoning of the type (cancer, "blackfoot disease," etc.) that prompted lowering the human drinking water standard from 0.05 to 0.01 mg As/L does not apparently occur in other animal species, as demonstrated by the ongoing search for an appropriate animal model to study the human condition. The mechanism(s) putatively involved in the pathogenesis of chronic damage in people, i.e. chemical attack by methylated AsIII metabolites on cellular macromolecules, do not appear to be relevant in livestock and wildlife. In domestic livestock, as opposed to people, most As is excreted via urine as DMAIII.33 This, together with the shorter observed half-life in these species, suggests that relatively little trivalent As escapes methylation and excretion to cause cancer. Chronic poisoning in livestock species involves mechanisms similar to acute poisoning and requires dosages very similar to acute poisoning.
Given the accumulating evidence that As is a human carcinogen, the question of residues arises. Can food animals consuming As from water accumulate dangerous amounts of As in edible tissues without themselves showing signs of toxicity? The literature to date suggests cattle, sheep, etc. eliminate As too quickly for this to be a concern, and a study completed in 2007 by the University of Minnesota89,90 failed to find any evidence of As accumulation in milk or edible tissues from dairy cattle watered from As-contaminated (140 μg/L) wells.
The threshold toxic dose in domestic ruminants appears to be between 1-2 mg/kg BW. This dose is in general agreement with the NRC25, which recommended 30-50 ppm dietary As as a maximum tolerated dose and with other reviews.88,91-93 It is quite distinct from the EU recommendation of 2 ppm dietary As, for which we have not been able to discover any justification. Sufficient quantitative data was not found to estimate a similar threshold for horses, but this dose is similar to that reported in another monogastric species (dogs)79, and previous reviews suggest horses are similar to cattle in sensitivity and/or less frequently affected than cattle under similar conditions.93,94 Therefore, it seems reasonable that limits safe for cattle should be adequate for horses. The very limited data in wild ruminants suggest they are similar to cattle in sensitivity. Therefore, our recommendations are based upon dosage data from cattle and sheep. Assuming neglible As in feedstuffs, 5 mg As/L in drinking water will provide the minimum toxic dose of 1 mg As/kg BW to grazing animals in warm weather. Obviously, if animals are receiving any As from forage or medications, less will be required to achieve a toxic dose. Although we were not able to find any significant studies of As in Wyoming forages, limited data from our laboratories suggest natural background concentrations seldom exceed a few ppm, except in areas contaminated by geothermal runoff.
Assuming a NOAEL of 0.5 mg/kg BW/day and allowing for these small forage concentrations, we recommend drinking water for livestock and wildlife not exceed 1 mg As/L