Results and Discussion
A partial summary of dietary and milk composition data is shown in Table 1. Whole roasted soybeans comprised 0 to about 15% of dietary dry matter. The high value is about the maximum inclusion rate recommended by most nutritionists. Average dietary copper was higher than recommended (10 to 20 ppm) concentrations. Dietary vitamin E also had a wide range and was much higher than NRC recommendations but similar to Ohio State University recommendations. Milk fat and protein concentrations were typical (one herd was comprised of Jersey cows). Mean concentrations of most milk constituents were typical of previously reported values. Concentrations of linoleic and linolenic acid were higher than average bulk tank milk. High correlations between dietary concentration of soybeans and concentrations of linoleic and linolenic acid were found (Figure 1). On average, milk fat from herds that were fed diets containing 15% roasted soybeans had about twice as much linoleic and linolenic acid as did milk fat from herds not fed whole soybeans.
| Mean | SD | Minimum | Maximum | |
| Dietary information (dry basis) | ||||
| Roasted soybeans, % | 4.2 | 3.9 | 0 | 15.3 |
| CP, % | 16.4 | 1.8 | 11.7 | 18.8 |
| NDF, % | 36.5 | 4.6 | 30.7 | 52.1 |
| Fatty acids, % | 4.6 | 0.9 | 2.7 | 6.1 |
| Copper, ppm | 27 | 9 | 14 | 59 |
| Vitamin E, IU/lb | 25 | 5 | 16 | 35 |
| Milk composition | ||||
| Fat, % | 3.83 | 0.27 | 3.28 | 4.70 |
| Crude protein, % | 3.25 | 0.17 | 2.99 | 3.82 |
| Milk fatty acids (% of total fatty acids) | ||||
| Linoleic acid | 3.6 | 0.8 | 2.2 | 6.3 |
| Linolenic acid | 0.7 | 0.2 | 0.3 | 1.3 |
| Copper, mg/L | 0.093 | 0.021 | 0.064 | 0.226 |
| β-carotene, mg/L | 0.03 | 0.03 | 0.01 | 0.29 |
| α-tocopherol, mg/L | 0.28 | 0.12 | 0.06 | 0.74 |
| Ascorbic acid, mg/L | 10.52 | 3.18 | 0 | 18.40 |
| Dehydroascorbic acid, mg/L | 2.41 | 1.86 | 0 | 9.39 |
| Flavor score1 | ||||
| 0 d of storage | 0.5 | 0.4 | 0 | 1.4 |
| 3 d of storage | 1.4 | 0.7 | 0.2 | 3.8 |
| 8 d of storage | 2.1 | 0.9 | 0.8 | 4.0 |
| 1Score; 0 = normal, 5 = intense oxidized flavor. | ||||
Over 8 days of refrigerated storage, milk fatty acid composition did not change (data not shown), however concentrations of vitamins E and C and β-carotene decreased markedly with 3 days of refrigerated storage (Table 2). The concentrations of vitamin E and β-carotene were reduced by about 20%, and vitamin C (ascorbic acid) was reduced by 70%. Vitamin E, vitamin C, and β-carotene are antioxidants and the rapid decrease in their concentrations increases the susceptibility of milk to become oxidized. However, concentrations of those antioxidants were not correlated with milk flavor suggesting that antioxidants when present in the concentrations we measured cannot eliminate oxidation when milk has high concentrations of polyunsaturated fatty acids.
| Storage time | ||
| 0 days | 3 days | |
| Vitamin E, mg/L | 0.28 | 0.22 |
| β-carotene, mg/L | 0.028 | 0.022 |
| Ascorbic acid, mg/L | 10.4 | 3.1 |
| DHAA, mg/L | 2.4 | 2.1 |
On day 0 (the day samples were collected), 92% of the samples had normal milk flavor scores (scores <1). After 8 days of storage, 17% of the samples had strong or intense oxidized flavor (score > 3). Milk factors that were most strongly positively correlated with milk flavor were concentrations of polyunsaturated fatty acids in milk fat and concentrations of dietary soybeans (Table 3). The concentrations of sixteen carbon fatty acids were negatively correlated with milk flavor but those relationships were probably indirect since concentrations of polyunsaturated fatty acids were negatively correlated with 16-carbon fatty acids.
| Correlation coefficient (r) | |
| Oleic acid | -0.34 |
| Linoleic acid (18:2) | 0.49 |
| Linolenic acid (18:3) | 0.55 |
| Total polyunsaturated | 0.50 |
| Polyunsaturated index | 0.53 |
| Roasted soybeans | 0.38 |
| 1Milk fatty acids expressed as percent of total milk fatty acids; polyunsaturated index = 18:2 + (18:3 x 2); roasted soybeans expressed as percent of total dietary dry matter. | |
Multiple regression was used to develop equations that could predict milk flavor score from milk composition data. Four different equations were derived that had similar accuracy. All equations included the concentrations of milk copper and dehydroascorbic acid (a breakdown product of vitamin C). Individual equations (Table 4) included concentrations of linolenic, linoleic, or total polyunsaturated fatty acids, or polyunsaturated index [linoleic + (linolenic x 2]. The polyunsaturated index is the concentration of polyunsaturated fatty acids weighted by the number of oxidizable chemical bonds in the molecule.
The regression equations make chemical sense. Based on the shape of the response surface, milk with high concentrations of copper (pro-oxidant) and high concentrations of polyunsaturated fatty acids (oxidizable substrate) were most likely to develop oxidized flavor. Milk with high concentrations of copper but low concentrations of polyunsaturated fatty acids (and vice versa) were much less likely to develop oxidized flavor. Dehydroascorbic acid is a breakdown product of vitamin C and the positive coefficient for dehydroascorbic acid in the equations suggest that increased breakdown of vitamin C is related to oxidized flavor.
| Fat | |||||||
| Model | Intercept | Type | Coeff. | Copper | Cu x fat | DHAA | R2 |
| 1 | 4.27 | 18:2 | 0.90 | -49.97 | 16.13 | 0.13 | 0.43 |
| 2 | 1.90 | 18:3 | 0.29 | -19.94 | 30.03 | 0.09 | 0.40 |
| 3 | 5.34 | PUFA | 0.80 | -63.33 | 13.71 | 0.12 | 0.44 |
| 4 | 5.22 | PI | 0.74 | -61.84 | 12.90 | 0.12 | 0.46 |
| 1Inter = intercept; Fat type; 18:2 = linoleic acid; 18:3 = linolenic acid; PUFA = polyunsaturated fatty acids (18:2 + 18:3 + conjugated linoleic acid); PI = polyunsaturated index (18:2 + (18:3 x 2)). All fatty acids expressed as percent of total milk fatty acids. Copper and dehydroascorbic acid (DHAA) expressed as mg/L. | |||||||