Bone analysis 126

A 3-point bending testing was used to determine the bone strength of the left femur. The bones were dissected from the hind legs and thawed at room temperature for approximately 14 hours (overnight). Once thawed the femur bones were weighed and bone length and diameter were measured using digital callipers. The breaking strength test was conducted on each bone, using an Instron materials testing machine (Model 3366, Instron, High Wycombe, Buck, UK) and a 3-point bending jig (Instron 5KN Flexure fixture 3- point bend). Using a 50 kgF load cell with a crosshead speed of 25mm/min and an attached anvil measuring 50 mm in length and 10 mm wide, force was applied to the midpoint of the same facial plane of each bone supported by two supports 40 mm apart. The bone breaking point was detected and force (kgF), strain and stress (kgcm2_{) were obtained using the Instron system}

4.3.8 Quantification of Immunoglobulin G

The concentration of immunoglobulin G (IgG) in colostrum and milk was assayed using specific pig- ELISA kits (Bethyl Laboratories Inc., Universal Biologicals, Cambridge, UK). The IgG was quantified according to the manufacturer’s instructions.

4.3.9 Statistical Analysis

All continuous response variables were modelled using linear mixed model methodology (REML). Binary and count variables were modelled using generalized linear mixed model methodology (GLM) with a binomial distribution and logit link function for the binary variables and a Poisson distribution and logarithm link function for the count variables. In all analyses, nursing sow and batch where included as random effects, while parity and treatment were included as fixed effects. For each response variable additional explanatory variables were fitted as fixed effects. A backwards elimination procedure was applied to these additional fixed effects for each response variable so that only variables that were significant at the (P<0.05) level remained in the final model in each case. All models were fitted using residual maximum Likelihood in the statistical software package GenStat (18th edition, VSN Internal Ltd, Hemel Hampstead, UK). If differences detected were significant, comparisons between groups were conducted with the fisher’s least significant difference test. Ordinal variables were fitted with the same linear mixed modelling strategy using a proportional odds model in the statistical package Stata (version 14.2, StataCorp LLC, 4905 Lakeway Drive, College Station, Texas 77845-4512, USA).

4.4 Results

4.4.1 Diet composition

In this study, the results from chemical analysis of the diets differed somewhat from the formulated analysis (Table 4.1). In particular, diet 3 had a lower crude protein (CP) value than the other diets. The amino acid content (AA) was also reduced. The diet 4 had lower crude fibre (CF) value than all other diets. Sub- samples of the diets were analysed and had consistently lower CP or CF than other diets. As the same batch of ingredients were used in the manufacturing of each diet and all four diets were made in parallel, the reduced CP and CF are unexplainable. The fatty acid composition (C4:0 to C22:6c) of the dietary treatments were analysed, with mean fatty acid values greater than 1.0 g/100g total fatty acids reported (Table 4.2). As intended, the different oil types changed the profile of fatty acids in the diets. The most pronounced differences were the presence of C20:5c and C22:6c, the increase in the proportion of C13:0, C16:0 as well as a decrease in C18:2c, when salmon oil was included in the diet compared with soya oil. As a result, diet 3 and 4 contained more saturated and n-3 fatty acids but had a lower n-6: n-3 fatty acid ratio than diet 1 and 2. Fatty acids below 1.0 g/100g total fatty acids accounted for approximately 2.0 % and 4.5 % of the fatty acid content of the soya and salmon oil diets, respectively. Although values of <1.0 g/100g total fatty acids are not reported they are included when applicable in the total saturated, monounsaturated fatty acid (MUFA), polyunsaturated fatty acid (PUFA), n-3 and n-6 calculations. Vitamin D3 content of the diets are mean values from 2

determinations extracted from bulked samples of finished feed. The gestation diets were formulated to contain either 2000 IU/kg (high) or 800 IU/kg (low)

vitamin D3 level but actual values were on average 2755 IU/kg in the high

vitamin D3 level diets and 1195 IU/kg in the low vitamin D3 level diets. The

vitamin D3 level in the gestation diets were greater than expected and may be

a result of contribution of vitamin D3 from raw materials, but the author is

confident this did not impact on the findings of the study. Actual DE and CP levels in the lactation diet, 17.2 MJ DE/kg and 18.8 g/kg CP, differed from expected levels. Actual Vitamin D3 level in the lactation diet was 2290 IU/kg.

Table 4. 2. Fatty acid composition (g/100g total fatty acids) of experimental diets fed during gestation

Gestation treatment

Diet number 1† 2 3 4

Oil Soya Soya Salmon Salmon Vitamin D31 High Low High Low

Fatty acid2 C13:0 38.38 37.62 41.41 42.97 C16:0 14.78 14.62 15.26 15.47 C18:0 2.67 2.79 2.29 2.22 C18:1c9 18.97 19.33 24.72 24.13 C18:1c11 1.76 1.71 2.32 2.30 C18:2c 53.45 53.30 36.52 38.10 C18:3cn3 5.94 5.99 5.70 5.70 C20:1c <1.00 <1.00 2.32 2.10 C20:5c <1.00 <1.00 1.91 1.75 C22:6c <1.00 <1.00 2.59 2.26 Total: Saturated3 _{57.08 56.25 61.66 63.23} MUFA4 _{21.39 21.71 30.76 29.84} PUFA5 _{59.82 59.57 48.67 49.59} n-36 _{6.21 6.14 11.21 10.64} n-67 _{53.62 53.44 37.46 38.96} n-6:n-38 _{8.63 8.70 3.34 3.66}

†Control diet fed day 0-29 of gestation prior to commencement of experimental feeding on day 30 of gestation

1_{High’-2000 IU/kg and ‘low’- 800 IU/kg vitamin D}

3

2_{Fatty acids are reported as g/100g of total fatty acids with a reporting limit of 1.0}

g/100g. Values presented are mean percentages of total lipid fraction from 2 determinations extracted from bulked samples of diet.

3_{Saturated- saturated fatty acids}

4_{MUFA- monounsaturated fatty acids}

5_{PUFA- polyunsaturated fatty acids}

6_{n-6- omega 6 fatty acids}

7_{n-3- omega 3 fatty acids}

4.4.2 Sow characteristics

There was no significant interaction between oil type and vitamin D3 level on

any sow performance measures (P>0.05). Average sow parity was 4.9 (SEM=0.2). Mean sow live-weight and back-fat depth on day 28 of gestation was 241.7 kg (SEM=2.15) and 19.6 mm (SEM=0.35), respectively. At day 107 of gestation average sow live-weight was 284.0 kg (SEM=3.60) and back-fat depth was 21.6 mm (SEM= 0.40). The average sow gestation length was 116.5 days (SEM=0.20) and lactation length averaged 26.5 days (SEM=0.20). At weaning mean sow live-weight was 260.3 kg (SEM=2.3) and back-fat depth was 19.8 mm (SEM=0.42). There was a significant effect of oil type on sow lactation feed intake (P<0.01) where sows offered diets containing soya oil ate on average 12.0 kg more during lactation than sows offered diets containing salmon oil. There was no effect of gestation dietary treatment on any other sow measures recorded (P>0.05).

4.4.3 Litter performance

There was no significant interaction between oil type and vitamin D3 level on

any litter performance measure recorded (P>0.05). Similarly, there was no significant direct effect of oil type or vitamin D3 level in sow gestation diets on

any litter performance measures recorded (P>0.05). Mean total born and number of piglets born alive was 14.9 (SEM=0.32) and 14.4 (SEM=0.31), respectively. As litters were standardised to 14 piglets within 24 hours of farrowing, average litter size after fostering was 13.7 (SEM=0.23). The mean number weaned was 11.4 (SEM=0.21). Mean total litter weight at birth was 20.7 kg (SEM=0.33) and coefficient of variation (CV) of litter birth weight was

21.2 % (SEM=1.14). The average daily gain (ADG) of litters was 2710 g/day (SEM=72.34) and mean pre-weaning mortality was 16.9 % (SEM=1.77). The average litter weight weaned was 90.95 kg (SEM=1.53) and CV of litter wean weight was 19.1 % (SEM=0.77). The number of mummified piglets and piglets fostered in and out were recorded but due to low numbers statistical analysis was not performed.