The Application of a Selenium Fertiliser for the Correction of Marginal Deficiencies in Grazing Sheep

INTRODUCTION Subclinical selenium (Se) deficiencies in sheep are often unpredictable and transient in nature. Associated subminimal concentrations of Se in forages, resulting in lower production levels, may remain undetected. Subclinical Se deficiencies and production responses to supplemen-tation were reported in the southern Cape coastal area 1 and the Kwa-Zulu Natal midlands


INTRODUCTION
Subclinical selenium (Se) deficiencies in sheep are often unpredictable and transient in nature.Associated subminimal concentrations of Se in forages, resulting in lower production levels, may remain undetected.Subclinical Se deficiencies and production responses to supplementation were reported in the southern Cape coastal area 1 and the Kwa-Zulu Natal midlands 21 .A number of methods can be used for supplementing Se to grazing ruminants.The use of mineral licks to provide Se is limited owing to low and variable intakes 4 .The monthly drenching of sheep with sodium selenite is cumbersome to perform under certain conditions 20 .Intraruminal Se pellets 10,11 or glass boluses 15 serve as a readily available source of Se, but these products are not commercially available in South Africa at present.
An alternative may be to provide Se by topdressing pastures with sodium selenate 8,22 .Because of the cost factor involved under extensive conditions, where a low stocking density is maintained, Se may be applied in strips to cover only 20 % of paddocks 4 .A later concept involved the application of poorly soluble barium selenate as a slow-release source of Se to pastures at a rate of 10 g Se per ha 23,25,26 .This product not only maintained elevated plasma Se levels for up to 3 years in sheep, but it also resulted in improved wool production and an increased live weight in sheep maintained on Se-fertilised paddocks compared to control animals on untreated paddocks 27 .
Against this background, we assessed the influence of a commercial Se fertiliser on the Se status of sheep on sites that were known to be marginally Se deficient 1 .

Location and treatments
The trials were conducted at the Outeniqua experimental farm (33°55' S, 22°25' E) near George in the southern Cape.The site experiences cool summers and mild winters, with a total annual long-term precipitation of 694 mm.The average monthly rainfall varies from 47.1 mm in July to 70.3 mm in October.The fine, sandy loam topsoil overlies structured clay, which is usually derived from granite and Table Mountain sandstone.It is acidic and low in trace minerals.Production responses of sheep as a result of oral or parenteral supplementation with copper (Cu) and selenium (Se) on the site have been reported previously 1,20 .
The commercial Se fertiliser used in the experiment was available in granular form.It consisted of a poorly soluble barium selenate core, coated with highly soluble sodium selenite.It has been shown that similar preparations allowed the sustained release of Se over a long period 25,26 .The fertiliser contained c. 10 % Se, and was administered at a level of 1 kg/ha in 2 trials.
Trial 1: Se fertiliser was administered to 3 kikuyu (Pennisetum clandestinum) paddocks ranging from 0.5 to 2.1 ha during May 1995.Three similar-sized paddocks without Se fertilisation were left as controls.All paddocks received equal quantities of supplemental irrigation.Pastures were grazed by 9-month-old SA mutton merino ram lambs, at an initial stocking density of c. 20 sheep per ha.Mature SA mutton merino wethers were used to balance numbers of sheep on differentsized paddocks.The efficiency of Se uptake on fertilised pastures was determined by the analysis of whole blood and organ Se concentrations.Heparinised blood samples (10 m ) from 5 designated ram lambs per paddock were obtained monthly (from May to September) by jugular venipuncture, and stored at -20 °C before analysis.During July and September, 5 ram lambs from each paddock were slaughtered to obtain liver and kidney samples.These samples were stored at -20 °C before analysis.The concentration of Se in whole blood, liver and kidney samples was determined by fluorometric procedures 13 .During 1996, these paddocks were similarly fertilised with Se.Another group of ram lambs utilised the paddocks at a somewhat lower stocking density of c. 15 sheep per ha.Sampling was conducted as described above, but monthly blood samples were obtained from June to November.Slaughter of 5 animals for liver and kidney samples took place in July, September and November.Towards the end of the trial, the flocks were disrupted by stock theft, and slaughter animals were available for only 2 of the 3 control paddocks in November 1996.
Trial 2: four of 7 paddocks were fertilised with Se in May 1995, as described previously.The remaining paddocks were left untreated as controls.Six of these paddocks were c. 0.7 ha in size, while the 7th was smaller (0.5 ha).An oat fodder crop was cultivated on all paddocks, which were utilised by groups of 9 to 15 pregnant SA mutton merino ewes.The ewes were introduced onto the paddocks at a fixed stocking rate of 20 ewes per ha in mid-June, just before lambing in July.After lambing, the ewes and their progeny grazed on the paddocks until docking at an average (±SD) age of 41 ± 8 days.At this stage, whole blood samples were obtained from individual ewes and lambs by jugular venipuncture.Lambs were weighed and identified with their mothers within 24 hours of birth.Live weight at docking and weaning (at an average age of 125 ± 9 days) were recorded.These data were used to determine lamb survival to weaning of individual lambs and the reproduction (number and weight of lambs weaned) of individual ewes.The blood samples were treated as described previously.The trial was repeated during June/August 1997, using the same paddocks.

Statistical analysis
The treatments were administered to paddocks, and not individual animals.It was thus impossible to treat individual animals as experimental units.Paddocks were therefore replicated and regarded as experimental units.Means were calculated for the respective paddocks, and used in the statistical analyses.The concentrations of Se in whole blood of the ram lambs in Trial 1 were analysed according to factorial designs 18 with treatment (Se fertilised or control) and month (May to September in 1995 and June to November in 1996) as factors.Liver and kidney Se concentrations were analysed according to a 2 (treatments) × 2 (slaugh-ter dates; July and September) experimental design in 1995.The corresponding analysis in 1996 was a 2 (treatments) × 3 (slaughter dates; July, September and November) factorial.
Whole blood Se concentrations of ewes and lambs in Trial 2 at docking were analysed as 2 (treatments) × 2 (years; 1995 or 1997) factorial designs.The Se treatment was represented by 4 replications per year, whereas 3 replications per year constituted the control treatment.Recordings obtained from individual ewes were used to calculate paddock means for reproductive performance.Ewe reproductive parameters included the mean percentage of lambs weaned and weight of lamb weaned per ewe available.As far as lambs were concerned, survival (lambs weaned expressed as a percentage of lambs born) was calculated on a within paddock basis.Paddock means for the continuous variables (live weight at birth, docking and weaning) were derived from least-squares analyses 6 .The fixed models fitted to the data included the effects of paddock, sex (ram or ewe), birth type (single or multiple) and the 2-factor interactions between these effects.This approach was used to ensure that paddock means were adjusted for unevenness in the distributions of sex and birth type, since individual paddock means were based on relatively small sample sizes within birth type classes.Age at docking and weaning were included as linear covariables, to adjust means for age differences between paddocks.Paddock means for the pro-duction traits were analysed according to a multi-factor analysis of variance, including years (1995 and 1997) and treatments (Se-fertilised or control) as factors.The 2-factor interaction was included in preliminary runs.Since it was not significant, only the main effects were retained in the final analysis.

Trial 1
No clinical symptoms of a Se deficiency were found in any of the trials.Whole blood Se concentrations of ram lambs were affected (P < 0.001) by treatment, month and their interaction in 1995 and 1996.The models fitted the data well, and accounted for a proportion of ≥ 0.89 of the available variation.During 1995, Se fertilisation of pasture markedly increased (P < 0.01) whole blood Se concentrations from approximately 80 ng/m to values exceeding 140 ng/m within a month (Fig. 1).After being introduced to the pasture for 4 months, blood Se concentrations increased to >300 ng/ml, followed by a decline to approximately 190 ng/m in September.Concentrations of whole blood Se of lambs on the control paddocks were maintained at approximately 80 ng/m for 3 months, before declining to <50 ng/m .
During 1996, whole blood Se concentrations of ram lambs on the fertilised paddocks were increased (P < 0.01) to >200 ng/m within 1 month of introduction (Fig. 2), with a maximum value of 290 ng Se per m whole blood in October.Whole blood Se concentrations of lambs on the control paddocks ranged from 76 ng/m to 87 ng/m .
During 1995, no interaction was observed between Se treatment and slaughter date as far as liver Se concentrations were concerned (Table 1).Liver Se concentrations of lambs grazing fertilised paddocks were 3.5 times as high (378 vs 1323 µg/kg) as that of lambs utilising the control paddocks in July, and 3.7 times as high (282 vs 1038 µg/kg) as that of lambs on control paddocks in September (P < 0.01).Time of slaughter and Se treatment interacted (P = 0.06) in 1996.Lambs on the fertilised paddocks maintained liver Se concentrations c. 4 times as high (477 vs 2051 µg/kg) as those on the control pasture in September (P < 0.01).The dif-ference between treatments was reduced in October and November, but remained significant (P < 0.01).

Trial 2
The model fitted to whole blood Se concentrations of the ewes and their lambs fitted the data well, accounting for >0.90 of the available variation.Whole blood Se concentrations of ewes and lambs were increased (P < 0.01) on the Se-fertilised paddocks relative to the control paddocks during 1995 and 1997 (Table 2).During 1997, the responses in both ewes and lambs were greater than in 1995.This resulted in an interaction (P <0.05) between Se treatment and year.
The model fitted to the ewe reproduction and lamb survival data had a poorer fit than those presented earlier, and accounted for respective proportions of 0.38 and 0.18 of the total variation in lamb survival and lambs weaned per ewe.The proportions of total variation accounted for in live weight traits ranged from 0.56 for birth weight to 0.11 in the case of weaning weight.Year effects were significant (P < 0.10) in some instances.No significant (P < 0.10) difference in ewe reproduction and lamb weights was observed between paddocks fertilised with Se and control paddocks.Absolute differences favoured the Se-treated paddocks for mean (±SE) lamb survival relative to the control paddocks (71.5 ± 4.6 % vs 62.2 ± 5.3 %, respectively; P = 0.21).

Blood and organ Se concentrations
Whole blood Se concentrations of the ram lambs grazing untreated control paddocks in Trial 1 were generally below 100 ng/m , regarded by some authorities as the threshold of marginal deficiencies 7 .Other workers suggested a threshold as low as 20 ng/m 9 .In subsequent work, a response to supplementation was observed at 76 ng Se/m whole blood in ewes rearing a lamb 12   Different superscripts in the same row denote significance (P < 0.01).Different superscripts in the same row denote significance (P < 0.01).
regarded as indicative of a marginal deficiency 7 .These findings suggest a marginal deficiency in Se on the experimental site, particularly during spring, as reported previously 1 .As far as the productive ewes were concerned, no conclusive evidence of a marginal Se deficiency was obtained, although the Se concentrations in the whole blood of control ewes and lambs approached 100 ng/m during 1997.Since the ewes were not exposed to the experimental paddocks for extended periods, this cannot be regarded as conclusive evidence of the lack of a subclinical deficiency on these paddocks.
The grazing of kikuyu pastures fertilised with Se clearly improved the Se status of ram lambs under these conditions, as reflected by increased whole blood and hepatic Se concentrations in Trial 1.A similar conclusion was evident in the case of the producing ewes in Trial 2. In the case of the ram lambs in Trial 1, this increase persisted for at least 6 months, making the product extremely suitable for rectifying the medium term Se status of grazing sheep.Increased Se concentrations in whole blood 1,11,12,14 , serum 12,17,28 , plasma 15 and liver 1,12,15,20 were commonly reported after oral or parenteral supplementation of Se under marginal conditions.It is conceded that the period over which the availability of Se was assessed in the present investigation was probably too short.When applied as fertiliser, sodium selenate was effective in increasing plasma Se concentrations of grazing sheep for up to 15 months 26 .In the case of slow-release barium selenate, maximum responses in plasma Se were only achieved during the 2nd and 3rd year following application 26 .This formulation was, in fact, effective in preventing subclinical Se deficiencies for at least 4 years after application 25 .

Production of ewes
Overall lamb production (defined as weight of lamb weaned per ewe mated) of ewes grazing pasture marginally deficient in Se was markedly improved by c. 50 % after supplementation 1 .The alleviation of a subclinical Se deficiency was found to result in improvements of all the components of overall reproduction rate, namely ewe fertility 5,29 , multiple birth rate 1,16 , lamb survival 10,14 as well as early growth of lambs and Angora goat kids 2,11,20,28 .The survival of beef calves was similarly improved after Se treatment 19 .Responses in lamb survival were often unpredictable and transient.Langlands et al. 10 , for instance, found that the administration of Se improved lamb survival relative to the control treatment at a high stocking rate (0.86 vs 0.74, respectively), while no effect was discernible at a low stocking rate (0.83 vs 0.84).The study of Segerson et al. 17 reported more stillbirths in control lambs compared to Se-treated contemporaries.Improvement in lamb survival could not be related to the thermoregulatory ability of lambs in another study 3 .
Improvement in ewe fertility and multiple birth rate was not expected in the present study, since supplementation only commenced shortly before lambing.The control ewes were furthermore not subjected to the unfertilised paddocks for extended periods of time.They could thus draw on Se reserves built up on the pastures that were utilised before the present trial.There was a suggestion of a better lamb survival on fertilised paddocks compared to control paddocks.Whether this is related to reported improvements in lamb survival following Se supplementation remains uncertain, and should be investigated in further production-orientated studies.

CONCLUSIONS
The medium-term Se status (6-8 months) of ram lambs grazing marginally deficient kikuyu paddocks was clearly enhanced by the application of a slowrelease Se fertiliser.A similar conclusion could be drawn with regard to the short-term Se status of ewes and lambs on an oat fodder crop, over a period of approximately 2 months.It must be conceded that the potential of the fertiliser to improve the long-term Se status (>12 months) of the pastures was not adequately assessed in this study.Viewed against the results of Whelan et al. 26 and Whelan and Barrow 25 , it is clear that products of a similar nature are able to provide sufficient Se to eliminate deficiencies for up to 4 years following application.Those studies, as well as that of Whelan 24 , suggest that fertilisation with BaSeO4 as a slow-release product provides a practical and safe way of Se supplementation to grazing ruminants for periods of up to 3 years after application.

Fig. 1 :
Fig. 1: Concentrations of whole blood Se in ram lambs grazing paddocks treated with Se fertiliser, or untreated control paddocks during 1995.Vertical bars denote standard errors.

Fig. 2 :
Fig. 2: Concentrations of whole blood Se in ram lambs grazing paddocks treated wih Se fertiliser, or untreated control paddocks during 1996.Vertical bars denote standard errors.

in ewes and lambs maintained on selenium-fertilised and control paddocks during 1995 and 1997. Based on a 2 × 2 factorial design with 4 replicates in the Se-fertilised treatment and 3 replicates in the control treatment.
Journal of the South African Veterinary Association 63: 148-155 22. Watkinson J H 1983 Prevention of selenium deficiency in grazing animals by annual topdressing of pasture with sodium selenate.New Zealand Veterinary Journal 31: 78-85 23.Whelan B R 1989 Uptake of selenium fertiliser by subterranean clover pasture in Western Australia.Australian Journal of Experimental Agriculture 29: 517-522 24.Whelan B R 1993 Effect of barium selenate fertiliser on the concentration of barium in pasture and sheep tissues.Journal of Agricultural and Food Chemistry 41: 768-770 25.Whelan B R, Barrow N J 1994 Slow-release selenium fertilisers to correct selenium deficiency in grazing sheep in Western Australia.Fertilizer Research 38: 183-188 26.Whelan B R, Peter D W, Barrow N J 1994 Selenium fertilisers for pastures grazed by sheep I. Selenium concentrations in whole blood and plasma.Australian Journal of Experimental Agriculture 45: 863-875 27.Whelan B R, Peter D W, Barrow N J 1994 Selenium fertilisers for pastures grazed by sheep II.Wool and live weight responses to selenium.Australian Journal of Experimental Agriculture 45: 877-887 28.Wichtel J J, Thompson K G, Graigie A L, Williamson N B 1996 Effects of selenium and iodine supplementation on the growth rate, mohair production, and thyroid status of Angora goat kids.New Zealand Journal of Agricultural Research 39: 111-115 29.Wilkins J F, Kilgour R J 1982 Production responses in selenium supplemented sheep in northern New South Wales. 1. Infertility of ewes and associated production.Australian Journal of Experimental Agriculture and Animal Husbandry 22: 24-29