Intake, water consumption, ruminal fermentation, and stress response of beef heifers fed after different lengths of delays in the daily feed delivery time

Four rumen-fistulated Holstein heifers (134 +/- 1 kg initial BW) were used in a 4 x 4 Latin square design to determine the effects of delaying daily feed delivery time on intake, ruminal fermentation, behavior, and stress response. Each 3-wk experimental period was preceded by 1 wk in which all animals were fed at 0800 h. Feed bunks were cleaned at 0745 h and feed offered at 0800 h (T0, no delay), 0900 (T1), 1000 (T2), and 1100 (T3) from d1 to 21 with measurements taken during wk 1 and 3. Heifers were able to see each other at all times. Concentrate and barley straw were offered in separate compartments of the feed bunks, once daily and for ad libitum intake. Ruminal pH and saliva cortisol concentrations were measured at 0, 4, 8, and 12 h postfeeding on d 3 and 17 of each experimental period. Fecal glucocorticoid metabolites were measured on d 17. Increasing length of delay in daily feed delivery time resulted in a quadratic response in concentrate DMI (low in T1 and T2; P = 0.002), whereas straw DMI was greatest in T1 and T3 (cubic P = 0.03). Treatments affected the distribution of DMI within the day with a linear decrease observed between 0800 and 1200 h but a linear increase during nighttimes (2000 to 0800 h), whereas T1 and T2 had reduced DMI between 1200 and 1600 h (quadratic P = 0.04). Water consumption (L/d) was not affected but decreased linearly when expressed as liters per kilogram of DMI (P = 0.01). Meal length was greatest and eating rate slowest in T1 and T2 (quadratic P <= 0.001). Size of the first meal after feed delivery was reduced in T1 on d 1 (cubic P = 0.05) and decreased linearly on d 2 (P = 0.01) after change. Concentrate eating and drinking time (shortest in T1) and straw eating time (longest in T1) followed a cubic trend (P = 0.02). Time spent lying down was shortest and ruminating in standing position longest in T1 and T2. Delay of feeding time resulted in greater daily maximum salivary cortisol concentration (quadratic P = 0.04), which was greatest at 0 h in T1 and at 12 h after feeding in T2 (P < 0.05). Daily mean fecal glucocorticoid metabolites were greatest in T1 and T3 (cubic P = 0.04). Ruminal pH showed a treatment effect at wk 1 because of increased values in T1 and T3 (cubic P = 0.01). Delaying feed delivery time was not detrimental for rumen function because a stress response was triggered, which led to reduced concentrate intake, eating rate, and size of first meal, and increased straw intake. Increased salivary cortisol suggests that animal welfare is compromised.

Aflatoxins and cyclopiazonic acid in feed and milk from dairy farms in Sao Paulo, Brazil

The occurrence of aflatoxins (AF) B(1), B(2), G(1), G(2) and cyclopiazonic acid (CPA) in feeds, and AFM(1) and CPA in milk was determined in dairy farms located in the northeastern region of Sao Paulo state, Brazil, between October 2005 and February 2006. AF and CPA determinations were performed by HPLC. AFB(1) was found in 42% of feed at levels or 1.0-26.4 mu g kg(-1) (mean: 7.1 +/- 7.2 mu g kg(-1)). The concentrations of AFM(1) in raw milk varied between 0.010 and 0.645 mu g l(-1) (mean: 0.104 +/- 0.138 mu g l(-1)). Only one sample was above the tolerance limit adopted in Brazil (0.50 mu g l(-1)) for AFM(1) in milk. Regarding CPA in feed, six (12%) samples showed concentrations of 12.5-1533 mu g kg(-1) (mean: 57.6 +/- 48.7 mu g kg(-1)). CPA was detected in only three milk samples (6%) at levels of 6.4, 8.8 and 9.1 mu g l(-1). Concentrations of aflatoxins and CPA in feed and milk were relatively low, although the high frequency of both mycotoxins indicates the necessity to continuously monitor dairy farms to prevent contamination of feed ingredients.

The influence of ARIMA-GARCH parameters in feed forward neural networks prediction

The objective of this article is to find out the influence of the parameters of the ARIMA-GARCH models in the prediction of artificial neural networks (ANN) of the feed forward type, trained with the Levenberg-Marquardt algorithm, through Monte Carlo simulations. The paper presents a study of the relationship between ANN performance and ARIMA-GARCH model parameters, i.e. the fact that depending on the stationarity and other parameters of the time series, the ANN structure should be selected differently. Neural networks have been widely used to predict time series and their capacity for dealing with non-linearities is a normally outstanding advantage. However, the values of the parameters of the models of generalized autoregressive conditional heteroscedasticity have an influence on ANN prediction performance. The combination of the values of the GARCH parameters with the ARIMA autoregressive terms also implies in ANN performance variation. Combining the parameters of the ARIMA-GARCH models and changing the ANN`s topologies, we used the Theil inequality coefficient to measure the prediction of the feed forward ANN.