Registration Dossier

Ecotoxicological information

Toxicity to other above-ground organisms

Currently viewing:

Administrative data

Link to relevant study record(s)

Description of key information

The data show that  all NOEC or LOECs at or below 100 mg/kg refer to studies with ruminants only. This finding is logical given the known effect of sulphides forming thiomolybdates in the anaerobic rumen, thereby reducing Cu absorption / Cu bioavailability in the blood stream in ruminants. Evidence shows that non-ruminant mammals are less sensitive, differences in dietary thresholds with cattle exceeding a factor 10. 
The differences in the physiological mechanism of Mo toxicity in ruminants and non-ruminant mammals is well understood. These findings advise to separate the hazard assessment for ruminants from that of non-ruminant mammals, and to consider cattle as the most sensitive of ruminant species.
Regarding non-ruminant mammals, the lowest NOEC value is 200 mg /kg dry weight. In terms of secondary poisoning, an assessment can be made considering, in combination, the BAF values with the toxicity in non-ruminant mammals. The modest BAF in combination with the low toxicity in non-ruminant mammals would suggest that the soil-earthworm-non-ruminant mammal pathway is unlikely to pose any risk to wildlife and that the generic PNECsoil will drive the risk assessment.
For ruminants, given the specific physiological mechanism rendering certain ruminants sensitive to Mo (molybdenosis), the particular pathway soil → grass → ruminants has been identified as requiring further attention.
Molybdenosis is a rare phenomenon occurring under specific conditions of low Cu:Mo ratios in the food (grass). An assessment is made using all available information, including field data. This pathway is thought to be of relevance at a local scale, where industrial emissions of Mo to air (deposited on grass/soil) may result in a disequilibrium of the Cu:Mo ratio in the ruminant diet.

Key value for chemical safety assessment

Long-term EC10, LC10 or NOEC for mammals:
200 mg/kg food

Additional information

The data show that all bounded NOEC or LOEC values at or below 100 mg/kg refer to studies with ruminants only. This finding is logical given the known effect of sulphides forming thiomolybdates in the anaerobic rumen, thereby reducing Cu absorption or Cu bioavailability in the blood stream in ruminants. The large species differences in susceptibility to molybdenum formed the basis to investigate mechanisms of Mo toxicity in mammals and birds. Mills and Davis (1987) reviewed these mechanisms and start with the observation that cattle are the most sensitive species to Mo intake, followed by sheep, both species are ruminants. Pigs, rats, rabbits, guinea pig and poultry (all non ruminants) are less sensitive, differences in dietary thresholds with cattle exceeding a factor 10. The differences in the physiological mechanism of Mo toxicity in ruminants and non-ruminant mammals and birds are discussed in some detail in the CSR. 

Briefly, the mode of action of Mo in most sensitive species such as cattle and sheep is due to the formation of thiomolybdates that inhibit Cu absorption or that act systemically with Cu bioavailability in the blood stream, both factors leading to Cu deficiency symptoms which can be cured by Cu supplementation.These reactions in the unique environment of the rumen explain Molybdenosis as a Mo and S induced hyprocuprosis.The disease is associated with Mo (“Molybdenosis”), because sulfate is available in sufficient amount in common forage, but the additional presence of Mo is critical. 

These findings advise to separate the hazard assessment for ruminants from that of non-ruminant mammals, and to consider cattle as the most sensitive of ruminant species.

Separate assessments for secondary poisoning are therefore made for non-ruminant mammals and ruminants.

Assessing the soil - earthworm - non-ruminant mammal pathway

Molybdenum concentration ranges in environmental matrices (soils, plants, earthworms, mammals and birds) are given in the section on bioaccumulation in the terrestrial compartment. The review of the data shows that Mo is not significantly concentrated from soil into plants or soil invertebrates, Bioconcentration Factors (BCF) or Bioaccumulation Factors (BAF<5) are typical and that there is no significant further concentration via the diet into mammals or birds when key organs such as kidney or liver have been examined (diet-tissues concentration ratio <10 and even <1 for muscle tissue).

Recent data has been determined on the concentrations of Mo in earthworms with Mo at near background concentrations, resulting in low ranges of BAF of 0.35 to 3.5 depending on the soils, and an average BAF of 1.5 (van Gestel, 2010). These BAFs generally decreased with increasing concentrations. This suggests thatbiomagnification of Mo, if any, is not significant in the terrestrial foodchain.

Toxicity data in non-ruminant mammals

Reliable experimental studies found in the literature on molybdenum toxicity to non-ruminant mammals are limited. The lowest NOEC for non-ruminants mammals was found at 200 mg Mo/kg in the rat study of Neilandset al. (1948). In this study, a dose of 500 mg Mo/kg diet reduced bodyweight gain in rats by ~ 35% although this study was not considered reliable.

The other studies showed toxicity of Mo intake to rabbits at 500 mg Mo/kg and to rats at 712 mg Mo/kg. Supporting information (experiments where only one Mo dose was used) show that rats exhibited symptoms of toxicity when a Mo dose was applied between 400 and 611l mg Mo/kg. The toxicity studies with swine showed that dietary Mo concentrations of 50 mg/kg did not affect growth (Dale et al., 1973) while a dose of 1500 mg/kg strikingly reduced growth by 45% (Standish et al., 1975) which was almost completely alleviated by an additional dose of 0.4%. sulphate in the diet. The NRC review on toxic thresholds of Mo in domestic animals refers to different swine feeding studies that failed to find toxicity between 100-1000 mg Mo/kg in the diet, suggesting that a bounded NOEC for swine is about 1000 mg Mo/kg (NRC, 1980).

The low BAF values combined with the high NOEC values for non-ruminants (all above 200 mg /kg dry weight)suggests that the soil-earthworm pathway is most unlikely to pose any risk to wildlife.

Assessing the soilàgrassàruminant pathway

All the existing data gathered on Mo toxicity to cattle as a single species is assessed since most of the studies have been conducted on this sensitive ruminant species.

All experimental and field studies on cattle are collated to evaluate dietary thresholds. In addition the nutritional factors explaining variability in dietary Mo thresholds is evaluated.

Dietary Cu:Mo thresholds for cattle

Experimental studies

In addition to 4 experimental studies with cattle conducted at several Mo doses, 12 experimental studies were found in which one single dose was given above the control. While useful, these latter tests can never be evaluated in isolation, since the dose-response relationship is not assured in a single study with one dose and no ‘bounded NOEC’ can be derived. Furthermore, in most of these studies, the control animals suffered from Cu deficiency.

In the selected studies the NOECs ranged from 20 to 100 mg Mo/kg. The lowest LOEC (Alopecia as an endpoint) was observed above the lowest NOEC of 20 mg Mo/kg found in the study of Vanderveen and Keener (1964). In the studies where one single molybdenum dose was applied, loss in bodyweight relative to the control was already found at a dose of 2 mg Mo/kg diet (Bremneret al., 1987). However, the background Cu concentration in the control diet was far below the average Cu grass concentration of EU samples (and Cu dietary requirements for cattle).

Several studies mention an effect on bodyweight at a Mo dose of 5 mg Mo/kg diet (Boyne and Arthur, 1986; Gengelbachet al., 1994; Humphrieset al., 1983; Phillippo and Humphries, 1987). In all these studies the background Mo in the control diet was also well below the dietary Cu concentrations recommended by the SCAN (2003). The similarity in all these studies was the 2 times lower value in Cu application than the average Cu grass concentration of8 mg/kg dry weight. In all other cattle studies where only one Mo dose was applied, effects of Mo addition were observed at concentrations ≥ 100 mg Mo/kg.

The Cu:Mo ratio in the diet of experimental cattle studies ranged between 0.02 and 2. The largest ratio of 2 was found in the study of Bremneret al. (1987) where an effect on bodyweight was already found with 2 mg Mo/kg and had a 4 mg Cu/kg diet.

Field data

The literature on field observations of molybdenosis was gathered. Field thresholds should be considered as adverse effects thresholds (i.e. a LOEC). Field thresholds are inherently uncertain since most of the studies are not ‘case controlled’, i.e. expert judgement was used to diagnose the disorder and attribute the symptoms to Mo. The field thresholds range from 2 to 930 mg Mo/kg.  The average dietary Mo thresholds ranged from 6-230 mg Mo/kg.

Merging field and experimental thresholds

Thevariability in thresholdsmay be partly explained by differences inanimal age,endpointsordietary factors. Risk of molybdenosis increases withreduced Cu supply,as shown in the experimental studies.

All experimental and field thresholds on Cu:Mo ratio are collated in Table 1. Observation shows adverse effects increase as the Cu:Mo ratio increases. The 90thpercentile of the data is 1.6 for experimental studies, > 1.30 for field observations (excluding the study of Smithet al., 1975) and 1.30 for the combination of the two. The study of Smithet al. (1975) was excluded as the enriched Zn concentration in the plants can have an antagonized effect on molybdenosis. The maximum Cu:Mo ratio in the experimental and field studies is 2.00.

Table 1: Summary statistics of dietary effects thresholds (Cu:Mo ratio in the diet for experimental studies; mean Cu:Mo ratio in the grass for field observations where effects were observed) for molydenosis in cattle; note that effects increase with decreasing Cu/Mo ratio, i.e. most sensitive animals are affected by largest Cu/Mo ratio; data were not selected for identifying confounding factors, age of test animals etc.

Cu:Mo ratio

N

Min

Mean

Max

90thpercentile

Experimental

12

0.02

0.47

2.00

1.64

Field (industrial&natural)

7

0.50

0.90

1.30

> 1.30

Combined (Experimental + Field)

19

0.02

0.61

2.00

1.30

**: observations of symptoms associated with molybdenosis were found on pastures with natural high Mo concentration in the grass and on pastures near industrial areas

The study of Smithet al. (1975) was excluded for this calculation as Zn could antagonize the effect of Mo in this study.

Comparing these findings with Cu:Mo ratios suggested in the literature

Thornton and Thums (2004) mentioned that cattle receiving diets with Cu:Mo ratios <1.0 are at high risk for molybdenosis, whereas diets with a Cu:Mo ratio between1.0-3.0 are indicative for a marginal risk.

Underwood and Suttle (1999) usedCu:Mo ratiosin the diet as a means of diagnosis, though stressed that they require flexible interpretation.

Erdmandet al. (1978) and Dollahiteet al. (1972) recommended a Cu:Mo ratio for cattle of about 6:1, whereas a ratio of less than 2:1 will most likely cause molybdenosis symptoms to develop.

Others (Webb and Atkinson, 1965; Alloway, 1973) believe that it is the concentration of Mo alone, and not the ratio of Cu to Mo, that is the common cause of molybdenosis.

A theoretical analysis suggests that the Cu:Mo ratio should be above 1 to avoid Mo induced Cu deficiency. The Mo-Cu-S complex formed in laboratory studies isCu1.6MoS4Xy(with X= Cl-or Br-and y = 1 or with X = SO42-and y = 0.5) (Laurie, 2000). The molar Cu:Mo ratio in this complex is 1.6. On a weight basis (MW Cu: 63.55; MW Mo: 95.94) this ratio becomes 1.0. Cattle receiving a diet with Cu:Mo ratio < 1.0 can thus be at high risk for molybdenosis.

Factors influencing the onset of molybdenosis

Molybdenosis occurs more readily atlower dietary Cu.Copper deficiency can be very detrimental for cattle grazing fresh forage but when the forage is dried as hay, no Cu deficiency is observed(Huberet al., 1971; Allaway, 1977). Copper bio-availability was evaluated in grazed pastures, dried grass, hay and silage by responses of plasma Cu in hypocupremic ewes by Suttle (1980). The bio-availability of Cu was significantly larger in cut grass and hay than in fresh grass and silage from the same field (e.g. the absorption of Cu from fresh grass ranged from 0.5 to 2.8% versus 7.2% for hay).

Individual experimental studies with cattle show that toxicity symptoms caused by an increased intake of molybdenum may be reduced or may totally disappear after the addition of copper to the diet (Clawson et al., 1972). Also in field studies, toxicity symptoms in cattle associated with an increased intake of molybdenum reduced or disappeared by addition of supplementary copper.

Bio-availability of Cu is also affected by thegenetics of ruminants. Differences between breeds were observed (Field, 1981; Wiener and Field, 1969). The Cu absorption of minerals from the diet in adult sheep ranged between 2 and 10% varying within breeds (Field, 1981). Some breeds of sheep exhibited Cu deficiency signs whereas other did not, when grazing on the same pastures in Scotland (Wiener and Field, 1969).

Total sulphur concentrations in forage typically range 1.5-4 g/kg. The Cu x Mo x S interaction is well reported forsheepand typically leads to higher Mo induced Cu deficiency in feeds that are higher in S. Full factorial designs have led to equations predicting the %Cu absorption in ewes as a function of Mo and S in the herbage or hay (Suttle 1991).

Forcattle, the Cu x Mo x S interaction is less conclusive. Several studies (Cunninghamet al. (1959 , Vanderveen and Keener (1964) clarify that the Cu in the diet is much more important than S in affecting the Cu status and it is speculated that the inorganic S in the control diet was already sufficient to reduce the Cu status in the Mo dosed animals.

Concluding, Cu:Mo ratios in the diet have been used as a means of diagnosis, although they require flexible interpretation given the influence of other parameters on the onset of molybdenosis. The Cu absorption is not only influenced by the presence of Mo and S but is also dependent on the type of grass (fresh or dried as hay) and the breed itself. Molybdenosis can be alleviated by Cu supplementation.

Dietary Cu:Mo thresholds for ruminants: a summary

Using all experimental and field thresholds on Cu:Mo ratios allows for a general picture of the onset of molybdenosis. The 90thpercentile of the Cu:Mo ratios is taken forward for further use in assessing the potential risk of molybdenosis at a regional scale. The 90thpercentile of Cu:Mo ratio predicts that the diet of ruminants should contain a Cu:Mo ratio of at least 1.3. It should be noted that this Cu:Mo ratioassumes 100% exposure via herbage. A refinement based on S content of forage is not possible for cattle due a lack of quantitative information on the Mo x S interaction for that species. Cattle is the most sensitive species, other ruminants that have been studied such as sheep, mule deer, goat, red deer are markedly less sensitive. Other herbivorous non-ruminants such as horses are do not display molybdenosis and are less sensitive to molybdenum. This means that the threshold for cattle can be taken forward for a risk assessment of ruminants in general.

The following uncertainty analysis should be taken into account :

Conclusion

In conclusion, there is sufficient scientific evidence to show that secondary poisoning is not an issue for molybdenum (low BAF in aquatic and terrestrial organisms and low toxicity in birds and mammals).

However, given the specific physiological mechanism rendering certain ruminants sensitive to Mo (molybdenosis), the particular secondary poisoning pathway (soilàgrassàruminants) has been identified as requiring further attention.

Molybdenosis is a rare phenomenon occurring under specific conditions of low Cu:Mo ratios in the food (grass). In terms of the REACH Chemical Safety Report, there is sufficient evidence to demonstrate that molybdenosis is not relevant for theregionalassessment. Therefore, theonlyexposurepathway of relevance is at the local scale,where industrial emissions of Mo to air (deposited on grass/soil) result in higher than background molybdenum in and on the food taken up by ruminants.

This specific assessment and the appropriate risk management measures are discussed in a separate section of the dossier as atargeted assessment on ruminants at the vicinity of industrial sites emitting molybdenum.


BIRDS

Molybdenum concentration ranges in environmental matrices (soils, plants, earthworms, mammals and birds) are given in the section on bioaccumulation in the terrestrial compartment. The review of the data shows that Mo is not significantly concentrated from soil into plants or soil invertebrates, Bioconcentration Factors (BCF) or Bioaccumulation Factors (BAF<5) are typical and that there is no significant further concentration via the diet into mammals or birds when key organs such as kidney or liver have been examined (diet-tissues concentration ratio <10 and even <1 for muscle tissue).

Recent data has been determined on the concentrations of Mo in earthworms with Mo at near background concentrations, resulting in low ranges of BAF of 0.35 to 3.5 depending on the soils, and an average BAF of 1.5 (van Gestel, 2010). These BAFs generally decreased with increasing concentrations. This suggests thatbiomagnification of Mo, if any, is not significant in the terrestrial foodchain.

Toxicity data in birds

Literature studies were reviewed. Two repeated dose experimental studies were found in the open literature.  Only 1 reliable feeding study with birds was found (Davieset al. 1960). In the study of Reidet al. (1956) unbounded NOECs were found for experimental studies with chickens and turkeys. The NOEC in the 28 day study of Davieset al. (1960) is 400 mg Mo/kg. 

Discussion

The soil-worm-bird food chain was used to assess if birds are at risk. The TGD recommended methods for secondary poisoning take into account that birds do not feed 100% of their time at a local soil and the recommendation is to estimate the exposure with 50% ‘regional’ and 50% ‘local’ food. The regional soil Mo concentration is about 1 mg/kg (median 90P of soils from EU-27 from GEMAS), the average and maximum soil-earthworm BAF values are about 1.5 and 3.4 respectively (see the chapter on bioaccumulation in the terrestrial compartment) and therefore the estimated regional Mo concentrations in earthworms are 1.5 x 1 = 1.5 (average BAF) or 3.4 x 1 = 3.4 (maximal BAF).

Assuming that the local soil is at 40 mg/kg (the maximum measured concentration of molybdenum from a local site which also corresponds to the generic PNECsoil for Mo), then the local Mo concentration in earthworms is 1.5 x 40=60 (average BAF) or 3.4 x 40 = 136 mg/kg (maximal BAF).

This leads to a dietary exposure at the maximum soil concentration of 0.5 x 1.5 + 0.5 x 60 = 31 mg/kg (average BAF) or 69 mg/kg (maximal BAF).

The lowest reliable NOEC value for birds is 400 mg/kg dw, i.e. 13-fold the average dietary Mo at the maximum local soil concentration, and 6-fold over the maximum local concentration combined with a maximum BAF. In addition, note that these earthworm BAF values may overestimate the environmental concentrations since the experimentral soils were freshly spiked for their determination, i.e. soils with relative high Mo bioavailability.

This assessment demonstrates that the soil-worm-bird pathway is not an important pathway for molybdenum toxicity and that deriving a PNECoral for this pathway is not relevant given that effects would start at more than 13 times higher than the highest Mo concentration measured in a local soil in Europe and that the generic PNECsoil will drive the assessment.