Publication Database

Validation of possible cow-side parameters for early detection of sickness in dairy cattle (2011)

Art

Hochschulschrift

Autor

Burfeind, Onno (WE 19)

Quelle

Berlin: Mensch und Buch Verlag, 2011 — 52 Seiten

ISBN: 978-3-86664-850-0

Sprache

Englisch

Verweise

URL (Volltext): https://refubium.fu-berlin.de/handle/fub188/176

Kontakt

Tierklinik für Fortpflanzung

Königsweg 65
Haus 27
14163 Berlin
+49 30 838 62618
fortpflanzungsklinik@vetmed.fu-berlin.de

Abstract / Zusammenfassung

An early detection of signs of sickness and consequently an early start of a therapy is beneficial for a successful treatment of diseases in dairy cattle. Due to a continuous increase in herd size there is less labour potential per cow and calf, respectively. Thus, the systematic evaluation of clinical parameters of animals at risk of disease on a regular basis might help to select animals for a clinical examination. The overall objective of this study was to validate cow-side parameters that might help to detect diseases in dairy cows and calves at an early stage. In the first study the variability of rectal temperatures in dairy cows considering different factors (intra- and inter-investigator repeatability, different thermometers, penetration depth into the cows´ rectum, and defecation) was validated. In a first experiment rectal temperature was measured multiple times per cow (n = 33) to determine intra-investigator repeatability. Repeated measures by a single investigator were consistent (39.5 ± 0.1ºC, coefficient of variance = 0.2%). Within the 10 measurements on the same cow, the maximum difference varied between 0.1 and 0.5ºC. In the second experiment rectal temperature was measured by 2 different observers in 38 cows to determine inter-investigator repeatability. Correlation between measures taken by 2 investigators was high (r = 0.98; P < 0.001) and mean difference was 0.1 ± 0.2ºC (P < 0.01). Data from these experiments demonstrate a high repeatability of rectal temperature measurements within and between investigator(s). However, for some cows, the difference between lowest and highest temperature was considerable (0.5ºC in 2 cows and 0.4ºC in 5 cows). In Experiment 3 the variation among different thermometers was tested in an in vitro and an in vivo phase. During the in vitro phase a water bath was adjusted in 1.0ºC increments from 35.0 to 42.0ºC and temperature measured with a validated thermometer and the 4 thermometers to be tested. During the in vitro phase correlation was high between all 4 thermometers and the validated thermometer (r = 0.99; P < 0.001) and mean differences were minor (0.0 to 0.1 ± 0.1ºC). During the in vivo phase rectal temperature was measured with all 4 thermometers in 37 cows by a single investigator. The measures from the 4 different thermometers were highly correlated (r = 0.94 to 0.96; P < 0.001). The mean difference between thermometers varied from 0.1 to 0.3ºC. Agreement was high when both thermometers compared had either a short or a long probe. Experiment 4 tested the influence of insertion depth. A single investigator measured rectal temperature in 33 cows, inserting the probe either 11.5 or 6.0 cm. Measures at the 2 penetration depths were highly correlated (r = 0.95; P < 0.001), but the result was 0.4 ± 0.2ºC higher when the probe was inserted deeper into the rectum (P < 0.001). Experiment 5 tested the effect of defecation on rectal temperature, measuring rectal temperature before and after defecation in 20 cows. Differences in temperature before and after defecation were minor. In 2 cows the measured temperature was 0.3ºC or higher before defecation, whereas in 3 cows it was 0.3ºC or higher after defecation. In the remaining 15 cows the difference before and after defecation was less than 0.3ºC. Overall, the results of this study indicate that some care is required in generalizing rectal measures of body temperature in dairy cows. To develop individual- standard fresh-cow monitoring programs at the herd level, rectal temperature should be measured with the same thermometer on the same penetration depth. The objectives of the second study were: 1) to determine repeatability (i.e. interobserver and intra-observer) of rumen fill scoring, 2) to study variation of rumen fill scores throughout the day in ad libitum fed cows, 3) to evaluate relationships between visual rumen fill scores and DMI and 4) to compare visually estimated rumen fill scores with exact measurements of the depth of the paralumbar fossa. Experiment 1 was conducted to determine inter-observer repeatability of rumen fill scoring and to study variation of rumen fill scores throughout the day in ad libitum fed cows. Three investigator independently scored rumen fill of 42 cows 3 times a day (0800, 1400, 1900 h). Experiment 2 consisted of 3 replicates (67, 70, 71 cows) employing 3 observers. Cows in each replicate were scored twice by the same observer to determine intra-observer repeatability. The intra-observer reliability for the 2 rumen fill scoring sessions showed moderate agreement (κw = 0.69; rS = 0.66, P < 0.001). Similarly, the inter-observer reliability for 2 independent observers showed reasonable agreement in rumen fill scores (κw = 0.68; rS = 0.71, P < 0.001). All 3 observers recorded higher rumen fill scores at 1900 h compared to 0800 h and 1400 h (Observer A: 0800 h: 3.1 ± 0.7, 1400 h: 3.4 ± 0.7, 1900 h: 3.6 ± 0.7; Observer B: 0800 h: 3.2 ± 0.8, 1400 h: 3.5 ± 0.8, 1900 h: 3.7 ± 0.7; Observer C: 0800 h: 2.8 ± 0.7, 1400 h: 3.1 ± 0.8, 1900 h: 3.2 ± 0.7; n = 210). Experiment 3 was conducted to evaluate relationships between within-cow changes in visual rumen fill score and changes in DMI in multiparous cows. Within-cow changes in visual rumen fill scores were compared to changes in DMI calculated for a 24-h interval. Additionally, data collected during Experiment 1 was used to assess the relationship between within-cow changes in visual rumen fill score and changes in DMI comparing differences between daylight hours and the night. I did observe a relationship between changes in visual rumen fill scores and DMI within cow on a 24-h basis (rS = 0.1, P = 0.09; n = 288) and comparing differences between day and night (rS = 0.68, P < 0.01; n = 257). In Experiment 4, variation in the objectively measured depth of the left paralumbar fossa within cow (intra-cow variation) was evaluated and the relationship between the depth of the left paralumbar fossa and rumen fill score determined. The variation in depth within the 5 classes of rumen fill scores (intra-score variation) was also determined. The average intracow variation in the depth of the paralumbar fossa was moderate (5.59 ± 0.9 cm; CV = 16%). On the same cow the range varied between 1.2 and 4.8 cm, measuring 10 times over a period of 70 ± 5 min. Variation in depth within each of the different rumen fill scores (intrascore variation) was also high. Spearman´s rank correlation (rS) between the depth of the paralumbar fossa and the rumen fill score was moderate (Observer A: rS = -0.64, P < 0.001; Observer B: rS = -0.60, P < 0.001). These data illustrate that rumen fill scores show moderate intra- and inter-observer repeatability. Moreover, these scores are moderately correlated with an objective measures of the depth of the left paralumbar fossa; a measure that changes considerably over a 70 min period. We did observe a reasonable relationship between changes in DMI and rumen fill scores within cow comparing differences between day and night, two periods with a significant difference in DMI, suggesting that future use of this measure on farms should be done across time within cow. Further research is needed to identify more accurate cow-side estimates of DMI, and to determine if changes in rumen fill scores can be used to identify cows at risk for disease. The objectives of the third study were: 1) to determine inter- investigator repeatability of rumination data collected via direct human observation in heifers and calves, 2) to determine the accuracy of the Hi-Tag rumination monitoring system in comparison to direct visual observation in heifers and calves considering different ages and, 3) to study whether sound of milk suckling in bottle fed pre-weaned calves interferes with the rumination monitoring system. In Experiment 1, 2 observers independently recorded rumination behavior from 20 animals via direct human observation for an observation period of 2 h each to determine the inter-investigator reliability. The rumination times were highly correlated (r = 0.99, P < 0.001, n = 20) and differences were minor (0 ± 2 min, P = 0.91). In Experiment 2, 6 groups consisting of 5 animals (Group 4 = 10 animals) of different ages were used to test the accuracy of the Hi-Tag rumination monitoring system in comparison to direct visual observation (Group 1: 25 ± 2 d, 64 ± 3 kg; Group 2: 42 ± 2 d, 80 ± 15 kg; Group 3: 62 ± 1 d, 90 ± 11 kg; Group 4: 95 ± 10 d, 118 ± 7 kg; Group 5: 185 ± 1 d, 207 ± 15 kg; Group 6: 282 ± 7 d, 342 ± 14 kg). Coefficients of correlation were highest for Group 3 (r = 0.89; P < 0.001) and Group 6 (Group 6: r = 0.88: P < 0.001), lower for Group 1 (r = 0.65; P = 0.009), Group 2 (r = 0.70; P = 0.004) and Group 5 (r = 0.72; P = 0.002), and lowest for Group 4 (r = 0.47: P = 0.009). The differences were lowest in Group 2 (0 ± 12 min; P = 0.77), Group 3 (2 ± 10 min; P = 0.60), and Group 6 (-4 ± 8 min; P = 0.05), moderate in Group 1 (-8 ± 10 min; P = 0.01) and Group 5 (-8 ± 14 min; P = 0.03) and highest in Group 4 (-12 ± 16 min; P < 0.001). Variation was low in Groups 6 (7.8 %) and 3 (11.8 %), but higher in Groups 1 (14.7 %), 2 (14.4 %), 4 (25.0 %), and 5 (22.8 %), respectively. In Experiment 3 I determined whether distortion of the Hi-Tag rumination monitoring system occurred when pre weaned calves suckled milk through a nipple. Each of 9 calves was fitted with a rumination logger just before milk feeding in the morning and again in the afternoon and the loggers removed immediately after the calves had finished suckling. They were stored in a quiet room for the rest of the interval and another 2 h to create a negative control interval. The coefficient of correlation between suckling time determined by direct observation and the estimates provided by the Hi-Tag system was low (r = -0.1; P = 0.70; n = 18). Only in 2 of the 18 2-h intervals (11.1%) did the Hi-Tag system record a 2 min observation; whereas, no observations were recorded during the other 16 intervals monitored during suckling or in the 18 intervals that served as negative controls. The Hi-Tag rumination monitoring system provides an accurate measure of rumination time in Holstein heifers older than 9 months. In animals younger than 9 months the variation of the data is high. However, an automatic detection of rumination time in heifers and calves might be a useful tool for detection of illness or studying rumen development in calves. Considering that suckling does not confound rumination time measures further research is encouraged to optimize the sound based detection system for rumination in calves. Overall, the three studies clearly evaluated simple and cow-side tests to detect cows at risk for disease. However, the user has to be critical regarding accuracy and repeatability. Measuring body temperature (objective parameter) in dairy cows has proven to provide reliable information. Furthermore, care is required to interpret visually estimated rumen fill scores (subjective parameter) and further research is needed to test, if rumen fill scores can be used to detect cows at risk for disease. Although the Hi-Tag rumination monitoring system does measure rumination time accurate in cows and heifers older than 9 months the variation of the data is high for younger animals.