Creating models for the prediction of colostrum quantity, quality, and immunoglobulin G yield in multiparous Jersey cows from performance in the previous lactation and environmental changes.

With multiparous Jersey cows, colostrum production seems to be variable. Due to this, we aimed to identify specific variables involved in colostrum production and quality. From 2021 to 2023, data from 28 US farms (415 multiparous Jersey cows) were used to investigate if colostrum yield, IgG concentration (g/L), and IgG yield (g) could be predicted by farm variables and transmitting abilities. With the data collected, multiple regression equations were developed to aid in predicting colostrum yield, IgG concentration, and IgG yield. Colostrum was weighed and sampled for IgG analysis. Dairy Herd Information (DHI), calving, diet, and management information data were compiled. Days below 5°C (D<), days above 23°C (D>), and days between 5 and 23°C (D) were recorded. We evaluated transmitting abilities for milk, fat, protein, and dollars; previous lactation milk yield, fat percent, fat yield, protein percent, protein yield, previous lactation somatic cell score, previous lactation days open, previous lactation days dry, previous lactation days in milk, and previous parity; and current lactation parity, days dry, and calving information, birth ordinal day, and latitude. Colostrum yield, IgG yield, and concentration had 1 added to correct for values = 0. After addition, values >0 were transformed to ln or log. Nontransformed variables were also used to develop the model. Variance inflation factor analysis was conducted, followed by backward elimination. The log colostrum yield model (R = 0.55; β in parentheses) included herd size (-0.0001), ordinal days (-0.001), ln ordinal days (0.07), latitude (-0.02), dry period length (0.004), D< (-0.005), D (-0.003), time to harvest (0.05), ln time to harvest (-0.35), IgG (-0.004), log IgG (0.46), feedings per day (0.06), ln pasture access (-0.13), and ln previous lactation days open (0.14). The model showed that previous lactation days open contributed the most toward increasing and latitude contributed the most toward decreasing colostrum yield. The IgG model (R = 0.21) included herd size (0.02), D> (0.38), ln time to harvest (-19.42), colostrum yield (-4.29), ln diet type (18.00), ln previous lactation fat percent (74.43), and previous parity (5.72). The model showed that previous lactation milkfat percent contributed the most toward increasing and time from parturition to colostrum harvest contributed the most toward decreasing colostrum IgG concentration. The log IgG yield model (R = 0.79) included ln ordinal days (0.03), time to harvest (-0.01), colostrum yield (-0.11), ln colostrum yield (1.20), ln pasture access (-0.09), ln previous lactation fat percent (0.53), and previous parity (0.02). The model showed that colostrum yield contributed the most toward increasing IgG yield, followed by previous lactation milkfat percentage. Pasture access contributed the most toward decreasing IgG yield, although the contribution was very small. These models were validated using 39 samples from 22 farms. Actual minus predicted colostrum yield and IgG concentration and yield were 0.89 kg, -21.10 g/L, and -65.15 g, respectively. These models indicate that dry period management and cow information can predict colostrum yield and IgG concentration and yield.