Vitamin B 12, in the blood of grazing cobaltdeficient

Transcription

1 New Zealand Journal of Agricultural Research SSN: (Print) (Online) Journal homepage: Vitamin B 12, in the blood of grazing cobaltdeficient sheep E. D. Andrews & B. J. Stephenson To cite this article: E. D. Andrews & B. J. Stephenson (1966) Vitamin B 12, in the blood of grazing cobalt-deficient sheep, New Zealand Journal of Agricultural Research, 9:3, , DO: / To link to this article: Published online: 14 Feb Submit your article to this journal Article views: 276 View related articles Citing articles: 16 View citing articles Full Terms & Conditions of access and use can be found at Download by: [ ] Date: 26 December 2017, At: 08:58

2 491 VTAMN B12 N THE BLOOD OF GRAZNG COBALT DEFCENT SHEEP By E. D. ANDREWS* AND B. J. STEPHENSON* (Received 3 June 1966) ABSTRACT Ewes grazing cobalt-deficient pastures were given cobaltic oxide pellets (cobalt "bullets") 23 weeks before lambing commenced. At intervals, blood samples, for vitamin B!2 assay, were taken from groups of untreated and treated ewes, and later, from groups or subgroups of untreated lambs from treated or untreated ewes, and from sub-groups of lambs given cobalt bullets directly. Also, livers and kidneys of lambs that died at or near to birth were assayed for vitamin B!2. Cobalt treatment of ewes resulted in markedly increased vitamin B!2 concentrations in the livers and kidneys of newborn lambs. However, although cobalt treatment of the ewes ameliorated, or delayed, the onset of cobalt deficiency disease in their lambs, the disease among the latter developed five weeks before weaning, that is, by the time the lambs were three to four months old. Treatment with cobalt bullets greatly increased vitamin Bn concentrations in the sera of ewes and lambs. At what appeared to be incipient stages of cobalt deficiency the disease was associated with mean serum Bn values of 0.26 m,ug per ml for ewes, and of 0.30 m,ug per ml for lambs. Thereafter, mean figures fluctuated widely from time to time, and at some subsequent stages exceeded those associated with incipient deficiency. However, at or towards the end of the respective experimental periods, when the effects of the disease had become marked in ewes and acute in lambs, mean Bn concentrations in sera in both cases fell to levels of <0.20 m,ug per m!. NTRODUCTON Several investigators (Hoekstra, Pope, and Phillips 1952; Kercher and Smith 1956; Dawbarn, Hine, and Smith 1957) have reported B12 values for blood from cobalt-deficient sheep. There is general agreement that, with the onset of the disease, blood B12 values fall markedly, and that levels are closely dependent upon the amount of cobalt ingested. Smith and Loosli (1957), in reviewing cobalt and vitamin B12 in ruminant nutrition, remarked that "t appears that for the first time we have a criterion, namely, blood vitamin B12 level, that will be of specific help in positively and quickly diagnosing a cobalt deficiency... " Also, Dawbarn et al. (oc. cit.) observed that "Symptoms of cobalt deficiency may be expected to supervene when the mean vitamin B12 * Wallaceville Animal Research Centre, Department of Agriculture, Private Bag, Wellington. N.Z. Jl agric. Res. 9:

3 492 Vitamin B12 in cobalt-deficient sheep activity in the blood plasma falls to about 0.2 mflg per ml." However, the previously mentioned studies were all carried out using penned animals consuming a semi-synthetic ration, presumably of constant cobalt content. t is reasonable to suppose that grazing animals would be subjected to greater stresses than are penned animals and that there would be some fluctuation of cobalt levels in pasture. These factors suggest that fluctuation in blood B 1 " levels might, in general, be greater for grazing animals than for penned animals. The present work was undertaken mainly for the purpose of exploring, further than hitherto, the value of assays of blood for vitamin B12, as a possible aid to the diagnosis of cobalt deficiency disease in grazing sheep. During the course of the work a number of newborn lambs died. Post-mortem examinations were carried out on those through the diagnostic services available at Wallaceville, and an opportunity was taken of observing the effect of cobalt supplementation of ewes on B J " concentrations in livers and kidneys of their lambs. The effect of supplementation of the ewe on the development of cobalt deficiency disease in the unweaned lamb was also examined. EXPERMENTAL On 22 February 1961, a reasonably even line of 60 two-tooth Romney ewes was obtained from the Department of Agriculture's Flock House Farm of nstruction, near Bulls, and transferred to known cobalt-deficient pastures at the Wallaceville Animal Research Centre's Farm at Kaitoke. The experiment started on 7 March 1961, when the ewes were divided at random into two groups of 30, bled from the jugular vein (for vitamin B12 assays), and weighed. Ewes in one group were each given a pellet containing 60% cobaltic oxide (cobalt "bullet," C.S..R.O. Formula C), together with a "grinder" (a 1 in. X 1 in. engineer's grub-screw). The other group served as a control. All sheep were each given a 11 oz dose of liquid "fine particle" phenothiazine on 27 March 1961, and again on 21 April. The ewes were mated, using a Romney ram, between 31 March and 30 May, and individual tupping dates were recorded. Weighing of the ewes continued at the intervals indicated in Fig. 1. Blood samples from all ewes were taken on the dates shown in Table 3 up to, and including, 27 July Difficulties due to failure of the vitamin B12 assay technique were encountered when the method was used for the June and July samples. Considerable time was necessary to trace and eliminate the probable source of error, and blood sampling was not resumed until 18 October. Similar difficulties arose for the November samples, but in that instance the probable cause of the failure was traced more quickly. From 18 October blood sampling of the ewes was restricted to 12 from each group (hereinafter termed the "selected" ewes), that had produced single lambs, and the results shown in Table 3 for the whole of the experimental period are for those animals only. To ensure maximal utilisation of feed the ewes were rotationally grazed over four different paddocks. However, it was necessary to give

4 E. D. ANDREWS AND B. J. STEPHENSON 493 some supplementary feed during winter, and for six weeks, from 19 June, chaffed meadow hay was made available. One ewe produced a dead lamb on 18 August. Otherwise, lambing commenced on 25 August, and continued until 30 September. Lambs were weighed at birth. and the birth dates were recorded. Except in one case all lambs that died at or near birth were taken for post-mortem examinations, and livers and kidneys were secured for vitamin B'2 assays. All surviving lambs were weighed on 18 October and again on 21 November. However, to avoid a possible bias due to uneven distribution of twins between groups, only the weights of 18 single lambs from untreated (control) ewes and 18 lambs from cobalt ewes are considered in this paper (see Fig. 2). Each group of 18 included the 12 lambs from each of the two groups (control or cobalt) of selected ewes. As part of another experiment (Millar and Andrews 1964) designed to investigate the retention of cobalt bullets by sheep, on 21 November all ewes in the cobalt group, and all lambs other than the 24 single lambs from the selected ewes, each received a cobalt bullet-on this occasion, one that had been made radio-active. So far as the present experiment is concerned this had the effect of dividing the 36 single lambs into four su b-groups, as follows:- A: No cobalt to ewes or lambs (12 lambs). A2: Cobalt to ewes; no cobalt to lambs (12 lambs). B : No cobalt to ewes; cobalt to lambs (six lambs). B2: Cobalt to ewes and lambs (six lambs). Lambs in sub-groups Al and A2 had been bled for B'2 estimations on 18 October and again on 21 November (Table 4). Further blood samples from those sub-groups, and from sub-groups B and B2, were taken thereafter at the dates shown in Table 4. The ewes were shorn on 16 December The lambs were weaned on 23 January 1962, and thereafter ewes and lambs were rotationally grazed as two separate mobs. By 23 February it was evident that surviving lambs from the 24 not given cobalt bullets (sub-groups Al and A2) would probably die if they were not treated. The survivors were, therefore, given a cobalt bullet, and the lamb experiment was discontinued from that date. Livers and kidneys from eight out of nine sub-group Al lambs that died and from the two lambs that succumbed in sub-group A2 were checked for vitamin B12 content. Although no control ewes had died by 23 February, it was evident at that time that many were in a weakened condition, and to save them each sheep in that group was given a single oral dose of cobalt sulphate, to supply 7 mg Co, on 2 March However, as indicated in Table 3, further blood samples were secured on 26 March and again on 8 May The 24 selected ewes were also bled on 11 June. However, results

5 494 Vitamin B 12 in cobalt-deficient sheep showed that, for all sheep examined, B,z concentrations in the sera had risen greatly, and inquiry revealed that two of the paddocks, through which the sheep had been rotationally grazed, had received a top-dressing of cobaltised sulphate in error between the May and June samplings. The ewe experiment was therefore discontinued as from 8 May. Cobalt concentrations in pasture samples taken from exprimntal paddocks on 21 April and on 18 October 1961 were checked by analysis. A sample of the chaffed meadow hay fed to the ewes during the v/inter of 1961, and sampled on 25 July, was also analysed for cobalt content. The method of analysis was that of McNaught (1948), and results are expressed as parts per million (p.p.m.) on a dry weight basis. Blood samples were allowed to clot, and assays were carried out as soon as possible on the sera, which were stored in a deep freezer. Vitamin B,z was extracted from 5 ml samples of serum according to an adaptation of the method of Ford (1953). n order to cope with the number of assays entailed. single samples of liver or kidney, or serum extracts, were used. For each sample five serial dilutions were prepared for assay according to the L'Actobacillus leichm'annii microbiological method of Krieger (1954), and results for liver and kidney are expressed here as flg per g (on a wet weight basis) and those for serum are expressed as mflg per m!. A test of reproducibility, carried out in triplicate on samples from bulked sera, gave figures of 1.52, 1.56, and 1.59 mflg per m!. Recoveries of cyanocobalamin, added to samples from the same source at the rate of 0.20 mflg per ml, were 0.18, 0.20, and 0.27 mflg per m!. However. it should be pointed out that B,z concentrations in liver and kidney are of an order of 10 3 times greater than those in serum, and, from experience gained during the course of the work, a high degree of reliability is not claimed for serum B'2 values given in this paper. Occasionally, obviously anomalous results were obtained for reasons that could only be surmised. n those cases, where the amount of serum available permitted, the assays were repeated. However, as indicated in Table 3, a number of results for sera from ewes was rejected, necessarily on a subjective basis. The inherent difficulties in microbiological assay techniques, particularly in those carried out on samples of low potency, are well recognised. n particular, according to a critical review by Shaw and Bessel (1960), L. leichmannii tube methods for vitamin B,z, althouh very sensitive, are probably those most susceptible to unpredictable variations. n this paper results for sera are presented and discussed subject to the foregoing reservations. RESULTS Cobalt concentrations in pasture samples taken in April and October were 0.04 p.p.m. and 0.05 p.p.m. respectively, that is, the values were consistent with cobalt deficiency disease in sheep. The cobalt level in the sample of meadow hay, fed during June and July, was 0.14 p.p.m., but, as indicated later, this normal concentration of the element apparently had little effect in delaying the onset of cobalt deficiency disease in untreated ewes.

6 E. D. ANDREWS AND B. J. STEPHENSON 495 Of the 30 cobalt ewes, all lambed. Of the 30 control ewes, 27 had lambs. Neo-natal mortalities were high due, probably, to very adverse weather during lambing. Of the 31 lambs from the cobalt ewes (29 single lambs and one set of twins) 10 died at or near birth. Of the 32 lambs from the control ewes (22 single lambs and five sets of twins), eight died. Mean birth weights for surviving lambs from ewes in each group were similar (l0.7b and lo.4lb for each of the two groups of 18 selected single lambs from the control and cobalt group of ewes, respectively). There was a small difference, which approached statistical significance, between the mean gestation period (expressed as days) for the 27 control ewes (148.3; S.E. ±0.4) and that for the 30 cobalt ewes (l47.2; S.E. ±0.4). Ewe 951 rejected its radio-active cobalt bullet sometime between 26 March and 8 May 1962 (i.e., about four or five months after administration), and ewe 986 rejected its bullet some time between 26 January and 23 February (i.e., about two or three months after administration). Radio-active bullets were retained by the remaining 10 selected ewes and by all the selected lambs over the periods of the present trials. Subsequent results of examinations on the retention of bullets are given elsewhere (Millar and Andrews loc. cit.). TABLE 1. Vitamin Bl2 in livers and kidlleys of lambs dying at or near birth Vitamin Bl') i (f.lg/g) - Ewe group Ewe No. 1 Liver Kidney _ Remarks Control D.15 Born dead. Dystokia. 978 D.31 ) D.18 Born dead. Dystokia. D.27 1 D. Born dead. Dystokia D.14 Found dead. Not suckled D.12 Lived 2 days. C/. septicum. ) 0.11 f 0.11 Lived 1 day. Not suckled D.ll 1 D.D6 Lived 1 day. Not suckled. 947 D.37 D.13 Lived day. Not suckled. E. coli. Means D.24.. Cobalt D.ll Born dead. Dystokia. 980 D , Born dead. Dystokia D.15 Found dead. Dystokia Born dead. Toxoplasma. 928 D Found dead. 944 D.6D D.26 Found dead. Suckled Lived 2 days. Suckled. 972 D.75 D.38 Born dead. Dystokia. 983 D.53 D.26 Born dead. Dystokia. Means 0.63 D.28 Bracketed figures are for twin lambs.

7 Vitamin B 12 in cobalt-deficient sheep Table 1 shows B," concentrations in the livers and kidneys of lambs that died at or near birth. The order shown for each group is the approximate order in which deaths occurred. t is evident that liver B12 values were almost invariably higher than the corresponding figures for kidney, and that, from inspection, differences between mean values for each organ are statistically highly significant. There is no evidence that liver/kidney Bu ratios were influenced by the time of death. infection, or whether or not the lamb had fed. Table 1 also shows that the mean B12 value for livers from lambs from cobalt-treated ewes was nearly three times higher than the corresponding figure for lambs from control ewes. The corresponding difference with respect to kidneys was about twofold. The difference, in each case, is statistically highly significant (P < 0.01). Results of assays of organs from 10 out of 11 of those lambs in sub-groups Alar A2, that died during the course of the lamb trial are summarised in Table 2. TABLE 2. Vitamin Bl Sub-group ill livers and kidneys of lambs dying dl/ring trial No. of lambs Liver Vitamin Bl (flg/g) Range Mean Range Kidney Mean At. No Co to ewe;; or lambs ,0.11. A2. Co to ewes, /10 Co to lambs Results shown in Table 2 for livers, and mean liver/kidney B'2 ratios, are indicative of cobalt deficiency disease of sheep (Andrews, Grant, and Stephenson 1963). Mean weight changes, for the 12 selected ewes in each group, are shown in Fig. 1. Statistical analyses of weight gains made by all the ewes suggested that incipient cobalt deficiency disease may have developed among untreated animals by mid-october (0.05<P<0.1), and showed that the disease had become firmly established by mid-november (P<O.O). At the start of the experiment the mean serum B2 value for all 60 ewes was 1.37 m,ug per m!, and, for the 24 selected sheep, was 1.40 mf.tg per ml. Because of failure of whole batches of samples during assay, and insufficiency of sample in a number of cases, to enable repeat determinations to be made, sets of figures for blood sera obtained from the ewes in June, July, and November, are incomplete. However, Table 3 shows

8 40.., CONTROL COBALT E. D. ANDREWS AND B. J. STEPHENSON :e! w :> Z "" :>:: u LAMLNG t----i t- :>:: :> iii DOSe) 7mg Co... "0, '0 M AM J J A SON 0 J F1M--.--,,----rM Fig..-Growth rates of control ewes and of cobalt-treated ewes. that B 1 " levels in the blood of untreated ewes had fallen sharply from July values by mid-october, that is, approximately coincidentally with the apparent onset of cobalt deficiency disease. There followed a further fall, from 0.26 m,ug per ml in October to 0.17 m,ug per ml in November. The subsequent rise to 0.31 m,ug per ml shown for December was for samples taken two days after shearing. Thereafter, except for the rise shown for the March samples (which followed the single dose of cobalt given 24 days previously, and which, as shown in Fig. 1, coincided with a temporary improvement in the mean weight of control animals), mean B12 figures again fell, eventually to < 0.20 m,ug per ml. Examination of individual weight changes showed that, from January 1962, each of the 12 selected cobalt-treated ewes made better gains than any of the control animals except one. The weight change pattern of this ewe (No. 982) was quite atypical of that for the control group as a whole, but was comparable to that for the cobalt group; in fact, at each weighing (from January 1962) the gain made by ewe No. 982 exceeded the mean gain for the cobalt-treated animals. Apparently the animal had not received a cobalt bullet in error, since its serum B12 values follow a pattern that is generally similar to the patterns for other control animals (Table 3), thus leading to.the conclusion that this particular ewe was resistant to exposure to the deficiency. The lamb from the ewe (lamb No. 932, Table 2) did, however, succumb to cobalt deficiency disease. Mean daily weight gains, and their standard errors, from birth until 18 October 1961 were 0.44±0.03lb and 0.50±0.02b, for the 18 single lambs from control ewes, and the 18 single lambs from cobalt ewes,

9 Group Ewe No. 11 April 1961 Control i i O.4ti i Uti ! _._... - Means 2.04 TABLE 3. Vitamin B12 (mjlg/ml) in sera of ewes Sampling dates - -"' ! i i 8 May June July : October November December January February March' May ! ()! * (0.17) 0.12 :;: (0.42) 0.26 O.S6 O.SO * (0.26) * 032 * (0.60) * (0.26) * * i (0.23) " UO 0.24 O.l (0.32) O.OS * (0.19) (0.22) * (0.16) (0.30) * (0.22) (0.28) 0.18 'C> 00 :s S t:l -"" -. ;:: <"'l '". ;::... '" ;::S. ';;:l i,.,, -

10 Cobalt 951(3) (1) Means (2) * * 4.74 * * * (, * * * * * * * * " No suitable figures available. (1) Bracketed B12 figures for samples taken 24 days after single 7 mg Co doses to control ewes. (2) Radio-active cobalt bullets given to cobalt group on 21 November (3) Ewe 951 rejected radio-active bullet between 26 March and 8 May (4) Ewe 986 rejected radio-active bullet between 26 January and 23 February ttl > Z :) ::c ttl :;;:: (j) ;J> Z :) l'=1... en - ttl "C ::: ttl Z (j) o z 'D \0

11 500 Vitamin B12 in cobalt-deficient sheep respectively. The difference, although not attaining significance (0.05 <P < 0.1), suggests that the lambs in the former group were incipiently deficient by mid-october..ll.::, lb is 20 i 10 o r 0 0 LAMBS FROM CONTROL EWES; NOT GVEN Co LAMBS FROM Co EWES ; NOT GVEN Co. o, (2) r... 0 LAMBS FROM CONTROL OR Co EWESiGVEN Co DRECTLY. OEA THS i THE NO. OCCURRNG DURNG EACH NTERVAL NDCATED AT TS BEGNNNG t" '" (2) t ,'... -'0... () () ( 7 ) BOCT NOV S DEC 26 JAM FEB Fig. 2.-Growth rates of lambs from control ewes and of lambs from cobalttreated ewes (to 21 November) and of cobalt-deficient lambs and of lambs given cobalt directly (from 21 November). Fig. 2 compares the mean weight gain made by the 18 single lambs from control ewes with that for the 18 single lambs from cobalt ewes, between 18 October and 21 November The means and their standard errors were respectively 6.9±0.8b and 13.6±0.8b, the 6.71b difference in favour of lambs from the cobalt group being significant (P<O.O). From 21 November 1961 to 26 January 1962 the performance of the 12 selected lambs from the cobalt-treated group of ewes (lamb

12 E. D. ANDREWS AND B. J. STEPHENSON 501 sub-group A2) was obviously better than that of lambs from the control ewes (iamb sub-group A) in terms of weight gain and death rate. However, between 21 November and 18 December 1961, means and their standard errors for weight gains made by sub-groups A2 and B2 were respectively 3.8±1.lb and 9.0±1.lb, the 5.21b response to cobalt being significant (P <0.01). Between 18 December 1961 and 23 February 1962 the mean weights of untreated surviving lambs, in sub-groups Al and A2, declined (Fig. 2). From Table 4 it is evident that, as expected, cobalt bullet treatment of lambs resulted in large increases in serum B12 values. However. an unexpected and inexplicable feature is that differences between ] anuary and February mean values for sera from sub-group B1 lambs and corresponding means for sera from sub-group B2 lambs reach statistical significance. The possibility of a major assay error is virtually excluded, since samples from both groups were assayed together. n conjunction with Fig. 2, Table 4 shows that definite cobalt deficiency disease, among sub-group A 1 lambs. was associated with a mean serum B12 value of 0.31 mflg per ml on 21 November, although the probability of incipient deficiency on 18 October, when the value was 0.30 mflg per ml, has been noted. However, the mean value rose appreciably in December, to 0.44 mflg per g. before falling to less than half that figure in January and February. For sub-group A2 incipient deficiency mayor may not have developed by 21 November (when the mean serum B12 figure was 0.32 mflg per ml). but the presence of the disease was apparent by 18 December when the mean serum B12 value was 0.36 mflg per m!. n contrast with the corresponding mean values for sub-group Al (which fell by more than half from the December to January samplings), the mean value for sub-group A2 was at approximately the same value in January as it was in December. but fell to less than half the preceding value in February. Thus, when the disease became more acute, it was associated with mean serum B12 levels of <0.20 mflg per ml, in January and February for survivors in sub-group A, and in February for those in sub-group A2. DSCUSSON Efficiency of treatments with cobalt bullets Although in practice the use of cobalt bullets has certain limitations (Millar and Andrews 1964), for the purposes of the present experiment it is evident that the bullets, when given to ewes or directly to lambs, were effective in preventing cobalt deficiency disease and in raising serum vitamin B12 values severalfold. However, the Bn levels for sera from ewe No. 986 (which rejected its radio-active bullet some time between 26 January and 23 February 1962) are of interest. From Table 3, between February and March, the level for this sheep fell sharply, and in March and May B12 values were appreciably lower than those for any other sheep in the cobalt group. A similar pattern appears to have occurred earlier, in October and November 1961, suggesting that ewe No. 986 may have rejected its previous bullet also, and that, in general, some individuals may be more prone to reject bullets than are others.

13 502 Vitamin 8 12 in cobalt-deficient sheep TABLE 4. Vitamin B12 (m/lg/ml) in sera of lambs Lamb sub-group Lamb No. Sampling dates October November December JanuarY!February \ : A. No Co to ewes or larr bs A2. Co to ewes. No Co to iambs B. No Co to ewes, Co to lambs B2. Co to ewes and lambs Means Mean \ * * ! * L * 0.23 * 0.22 * * 0.11 * * * Means Means ; * No sample, lamb dead.

14 E. D. ANDREWS AND B. J. STEPHENSON 503 Liver / kidney B 12 ratios: placental transfer of vitamin B 12 Liver /kidney B12 ratios for newborn lambs (Table 1) are in sharp contrast with those indicated in Table 2 for older cobalt-deficient lambs. n other work (Andrews and Hart 1962) it was found that for weaned lambs liver/kidney B'2 ratios were < 1 in the case of those that were cobalt-deficient, but were > 1 in the case of those that were normal. However, that finding need not, of course, necessarily apply to foetal or newborn lambs, in which the uro-genital system would not be fully functional. Except for the twin lambs from ewe No. 937 (Table ), liver B'2 values for lambs from untreated ewes would be considered normal for older sheep (Andrews, Hart, and Stephenson 1960), but again this need not necessarily apply to newborn animals. However, since the mean birth weight of lambs from cobalt ewes was actually slightly less than that for lambs from control ewes, it is reasonable to suppose that, as a group, the latter lambs were not seriously, if at all, affected by cobalt deficiency disease at birth. O'Halloran and Skerman (1961) observed that "Analysis of livers from lambs that were born dead, or died before sucking, indicated that lambs from (cobalt) treated ewes had higher reserves of vitamin B'2 at birth." However, their results were too few to allow definite conclusions to be drawn. Results in Table 2 show that cobalt supplementation of the ewe has considerable effect in increasing vitamin B'2 storage in the liver and kidney of the newborn lamb, and indicates that the vitamin is readily transferred via the placenta. The effect of treatment of the ewe on the development of cobalt deficiency in the unweaned lamb O'HaHoran and Skerman (loc. cit.) observed that cobalt treatment of ewes grazing cobalt-deficient pastures resulted in better growth of the lambs up until the time their final weightings were made, which was when the youngest lamb was 12 weeks and the oldest lamb 18 weeks of age. However, they had no groups of lambs that had been treated directly with cobalt, for comparison. Results in Fig. 2 confirm that cobalt treatment of the ewes will result in better growth of unweaned lambs. This can be accounted for by higher reserves of vitamin B'2 in the livers of lambs from treated ewes, and increased amount of B'2 in the ewes' milk (Harper, Richard, and Collins 1951; Moinuddin, Pope, Phillips, and Bohstedt 1953; Hart and Andrews 1959; O'Halloran and Skerman lac. cit.). However, cobalt deficiency among lambs from treated ewes had supervened by 18 December 1961, that is, five weeks before weaning, and when the ages of the lambs were between 11 and 17 weeks. Similar results had been obtained in a previous (unpublished) experiment carried out by the writers during the season. Hence it is concluded that cobalt treatment of the ewe will ameliorate, but not necessarily prevent, the development of cobalt deficiency disease in the lamb prior to weaning.

15 504 Vitamin B12 in cohalt-deficient sheep Serum B12 in relation to cobalt deficiency disease Consideration of the growth curve of ambs from cobalt-treated ewes from 18 October to 21 November and that for sub-group B2 thereafter (Fig. 2) suggests (but does not firmly establish) that lambs from cobalttreated ewes were not cobalt-deficient on 18 October, that is when mean serum B12 values were 0.52 mp,g per m. From the evidence, and for the purpose of this discussion, all other mean serum B12 values shown for ambs not given cobalt bullets are considered to be associated with cobalt deficiency disease. Similarly, for control ewes all mean values from 18 October 1961 inclusively (possibly excepting those for 26 March 1962) are considered to be associated with cobalt deficiency disease. 1mpressions from the data are that cobalt deficiency disease, in its earlier stages, may be associated with serum B," levels that are higher for unweaned lambs than they are for older sheep, and that, at least for the former, increasing unthriftiness may not necessarily be accompanied by progressive falls in serum B'2' but rather, by a precipitous fall once some threshold or critical condition has been reached. Present results may be compared with those found by other investigators under somewhat different experimental conditions. For sheep "in early stages of cobalt deficiency," Hoekstra et al. (loc. cit.) reported a mean vitamin B,"-activity in whole blood of 0.47 mp,g per ml. Kercher and Smith (loc. cit.), apparently working with whole blood (this is not specifically stated), found a mean value of 0.11 mp,g per ml for lambs in which cobalt deficiency disease was well established. However, perusal of their lamb weight data, obtained during a pre-experimental period when the deficiency was developing, suggests that the disease might have been affecting weight changes when mean B'2 levels were 0.30 mp,g per m. Working, apparently, with adult sheep, Dawbarn et al. (loc. cit.) found that vitamin B 12 -activity in blood plasma from two groups of 10 sheep grazing normal pastures was 4.4 and 6.1 mp,g per ml respectively. Five weeks later the sheep were confined to pens and fed a cobalt-deficient ration. After 103 days the mean B'2 figures had fallen to one tenth, or less, of the original values, but at that stage "the sheep showed no outward signs of the deficiency." After 210 days mean daily food consumption had fallen to 87 % of its highest value, but no B12 figures were given for that time. After 245 days, when mean daily food consumption had fallen by one third, mean B'2 values for the two groups were respectively and mp,g per ml, and it was concluded that symptoms of cobalt deficiency may be expected when the mean value falls to about 0.20 mp,g per ml. However, it is to be noted that no B'2 figures were available for the 20-week period between the 103rd day (when the mean B'2 value for one group was 0.39 mp,g per m!, and for the other, 0.47 mp,g per ml) and the 245th day, and perusal of all the data suggests that the critical B'2 level might have been closer to 0.30 m,ug per ml than to 0.20 mp,g per ml. Dawbarn et al. (loc. cit.), in discussing their vitamin B'2 results, remarked on the considerable variation found among individuals. Perusal of the results for one of their groups of 10 cobalt-deficient sheep shows

16 E. D. ANDREWS AND B. J. STEPHENSON 505 that, in approximate terms, there was a twofold difference between the lowest and the highest values, and for the other group there was a threefold difference. Neglecting those for 26 March, present results for cobalt-deficient ewes show that from 18 October there are threefold differences between lowest and highest values for samples taken on the same date, except for those taken on 26 January, where the atypically high value of 0.50 mf.-lg per ml (assay not repeated) results in a fivefold difference. Similarly, B12 results for lambs not directly supplemented with cobalt (sub-groups Al and A2) show twofold or threefold differences for different samples taken on the same date, except those taken on 26 January for sub-group A2 lambs, where the high figure of 0.88 m,ug per ml (assay repeated, and result confirmed) results in a sixfold difference. A tentative conclusion, therefore, is that at anyone time the variability in blood B," values among individual grazing sheep differs little from that for individual penned sheep and that for groups of cobalt-deficient animals a twofold to threefold difference between lowest and highest values for serum samples taken at the same time can be expected. From the viewpoint of interpretation of results an important source of variation resides in the fluctuations in mean values shown at successive times after establishment of cobalt deficiency conditions among groups. ft is of incidental interest that the appreciable though temporary rises in mean values shown for the cobalt-deficient group of ewes, and a1so for their lambs, on 18 December, followed shearing of the ewes two days previously. Elsewhere (Andrews, Grant, and Stephenson loco cit.) it has been remarked that "the physiological effect of shearing may in some way temporarily relieve the restriction on appetite imposed by a deficiency of vitamin B,"." However, whether the postulated effect on appetite would be sufficient to account, directly in the case of the ewes, and indirectly in the case of the lambs, for the rises in serum B," found, is questionable. All that can be said is that, for the ewes, presumed continuous cobalt deficiency disease was associated with mean values of both 0.31 mflg per ml and 0.16 mf.-lg per ml a twofold difference, and for lambs the disease was associated with mean values as high as 0.44 mflg per ml (sub-group A) and as low as 0.16 mflg per g a threefold difference. t is evident that diagnostic criteria should endeavour to take into consideration not only variability among individuals at the same time, but also variability of a group of individuals at different times. Tn considering detailed diagnostic criteria for sheep, the data for unweaned lambs (for practical purposes those less than five months of age) are somewhat less than satisfactory. However, indications are that, for samples from 10 or more animals, mean levels of <0.20 mflg per g are almost certainly indicative of cobalt deficiency disease, levels of <0.35 mflg per g are probably indicative of deficiency, and levels of >0.55 mflg per g are probably indicative of cobalt sufficiency. Unfortunately, because of serum B'2 data missed between July and October 1961, it is not possible, from present results, to suggest a useful mean level above which cobalt deficiency would not be expected to occur in ewes. However, in another experiment with which two of the writers were associated (Andrews, Stephenson, saacs, and Register in prepara-

17 506 Vitamin B12 in cobalt-deficient sheep tion), it was found that for young sheep approximately nine months old, which had received sufficient cobalt to maintain them apparently free from cobalt deficiency disease, the mean serum B'2 was 0.43 m,ug per ml The proposed diagnostic criteria which follow are based on the assumption that the findings for those animals apply to weaned lambs generally, and to adult sheep also. n view of the limitations of the assay method for serum discussed earlier and amplified later in this paper, indications are that about 10 or more animals, from each flock or group, should be sampled. On that basis, and from the foregoing discussion, the following criteria for sheep, five months of age or older, are suggested:- Mean serum B12 (,ug per m!) < >0.45 Cobalt status Definitely deficient. Probably deficient. ndefinite. NormaL t is not proposed to detail here the various factors that may render the results of a serum B'2 assay unreliable. t can, however, be said that, in our experience, failure of an assay may, in general arise in two ways. n most or all of the assay samples there may be little or no growth of the micro-organism. Errors of this sort appear, usually, to arise from improperly washed glassware, are not necessarily restricted to serum samples, and, of course, are readily detected. Alternatively, there may be contamination (bacterial or chemical) of unknown origin, of a single sample or, in rarer instances, of a batch of samples, at some stage prior to the setting up of each set of five serial dilutions, and leading to a spuriously high result. Replication of sample mayor may not facilitate detection of such errors, depending upon the stage at which contamination occurs. Clearly, the importance of undetected errors of this kind would depend upon the number of samples available from the flock or group, the magnitude of the error or errors, and the number of samples affected. t is apparent that because of the very great difference in vitamin B'2 potency between serum and liver, contamination of samples of the former is likely to have much more serious consequences. Also, it is likely that true B'2 concentrations in the blood of the animal are more susceptible to minor fluctuations in cobalt utilisation, or to extraneous factors such as liver damage, than are concentrations of the vitamin in the liver. Probably for those reasons, and perhaps for others at present not fully understood, our experience has been that results for liver can be accepted with considerably more confidence than can those for serum, and hence liver samples are preferred for diagnostic purposes. However, serum examinations would have value under circumstances in which liver samples were unavailable, and in investigations in which it was desired to follow the course of cobalt deficiency disease in groups of experimental sheep.

18 E. D. ANDREWS AND B. 1. STEPHENSON 507 ACKNOWLEDGMENTS The writers are indebted to: Miss Joan C. Kater, B.V.Sc., for postmortem reports on lambs; Mr F. K. Calverley, for the collection of blood samples; Dr L. 1. Hart, who advised on the application of the microbiological assay technique to sh:ep; and Mr G. Locke, for care of the experimental animals. They also thank Mr A. W. Barkus for the draughting of Figs. and 2. REFERENCES ANDREWS, E. D.; GRAt:n, A. B.; STEPHENSON, B : Wcight rcsponses of sheep to cobalt and selenium in relation to vitamin B\2 and selenium concentration in liver and kidney. N.Z. J agric. Res. 7: ANDREWS, E. D.; HART, L : A comparison of vitamin B" concentrations in livers and kidneys from cobalt-deficient lambs. hid. 5: ANDREWS, E. D.; HART, L..; STEPHENSON, B. J. 1960: Vitamin B" and cobalt in liver from grazing cobalt-deficient lambs and from olhers given various cobalt supplements. bid. 3: DAWlARN, MARY c.; HNE, DENSE C.; SMTH, JEAN 1957: The determination of vitamin B,,-activity in the organs and excreta of sheep. 5. The elrect of cjbalt deficiency on the vitamin B" content of blood plasma. Allst. J. expo Bioi. med. Sci. 35: FORD, J. E..1953: The microbiological assay of vitamin B12. The specificity of the requirement of Ochromonas malhamcllsis for cyanocobalamin. Br. J. Nutr. 7: HARPER, H. E.; RCHARD, R. M.; COLLNS, R. A. 1951: The influence of dietary cobalt upon the vitamin B" content of ewe's milk. Archs Biocllcm. B,'ophys. 31: BART, L..; ANDREWS, E. D. 1959: Effect of cobalt oxide pellets on the vitamin B12 content of ewe's milk. Nature, Lond. 184: HOEKSTRA, W.O.; POPE, A. L.; PHLLPS, P. H. 1952: Synthesis of certain B vitar;1:ns in the cobalt-deficient sheep with special reference to vitamin B". J. Nufr. 48: KERCHER, C. J.; SMTH, S. E. 1956: The synthesis of vitamin B" after oral and parenteral administration of cobalt to cobalt-deficient sheep. J. Anim. Sci. 15: KREGER, C. H. 1954: Report on vitamin B" microbiological method. J. Ass. off. agric. Chcm. 37: McNAUGHT, K : Spectrophotometric determination of cobalt in pastures and animal tissues. N.Z. Jl Sci. Tee/llo/. 30A: MLLAR, K. R.; ANDREWS, E. D. 1964: A method of preparing and detecting radio-active cobaltic oxide pellets and an assessment of their retention by 5heep. N.Z. vct. J. 12: MONUDDN, M.; POPE, A. L.; PHLLPS, P. H.; BOHSTEDT, G. 1953: The effect of ration in the riboflavin, niacin and vitamin B12 concentration of the blood and milk of sheep. J. Anim. Sci. 12: O'HALLORAN, M. W.; SKERMAN, K. D. 1961: The effect of treating ewes during pregnancy with cobaltic oxide pellets on the vitamin B12 concentration and chemical composition of colostrum and milk and on lamb growth. Br. J. Nufr. 15: SHAW, W. H. C.; BESSEL, CHRSTNE : The determination of vitamin B12. A critical review. Analyst, Land. 85: SM TH, S. E.; LOOSL, J. K. 1957: Cobalt and vitamin B\2 in ruminant nutrition: a review J. Dairy Sci. 40: