Estrus Detection in a Dairy Herd Using an Electronic Nose by Direct Sampling on the Perineal Region

Simple Summary Fertility a very important field for dairy farms and directly affects their economic success. Therefore, early and accurate estrus detection is crucial, particularly for farms using artificial insemination. During the last decades, several automated sensor-based technologies for estrus detection have been developed. Nevertheless, accurate estrus detection still remains a challenge. In this study, the capacity of electronic nose (EN) technology (MENT-EGAS prototype), which was based on 10 non-specified chemical metal-oxide sensors to detect estrus by the direct sampling of odor from the perineal headspace in Holstein dairy cows, was assessed. Principal component analyses (PCA) were applied and identified high discrimination between proestrus and estrus, and between estrus and metestrus in cycling cows. Additionally, high discrimination amongst estrus in cycling cows and pregnant cows was perceived. Based on these findings, we show for the first time that it is possible to recognize estrus accurately in dairy cattle by direct sampling on the perineal headspace using an EN device during milking. In the future, MENT-EGAS technology could be routinely used on dairy cattle farms as a non-invasive, accurate method for estrus detection. Estrus detection is very important for the profitability of dairy herds. Different automatic systems for estrus detection have been developed over the last decades. Our study aimed to assess the ability of the electronic nose (EN) MENT-EGAS prototype to detect estrus, based on odor release from the perineal headspace in dairy cattle by direct sampling. The study was performed in an Italian dairy farm using 35 multiparous Holstein-Friesian cows. The cows were divided into three groups: group I included 10 lactating 5-month pregnant cows, group II included 19 lactating cycling cows, and group III included 6 cows that were artificially inseminated 18 days before the trial. Odors from the perineal headspace were collected using the MENT-EGAS prototype. In group I, odors were collected once a day for 5 consecutive days. In group II, odors were collected twice daily from day 18 until day 1 of the reproductive cycle. In group III, odors were also collected twice daily from the presumable day 18 of gestation until day 22. Principal component analyses (PCA) of the perineal headspace samples were performed. PCA in group I revealed no significant discrimination. PCA in group II revealed clear discrimination between proestrus and estrus, and between estrus and metestrus but no significant discrimination was obtained between proestrus and metestrus. PCA in group III revealed that in four cows the results were similar to group I and in two cows the results were similar to group II. On day 40 of the presumable pregnancy, the ultrasound examination revealed that only the four cows were pregnant and the other two cows were regularly cycling. On the basis of our findings, we conclude that it is possible to accurately detect estrus in dairy cattle from directly collected odor samples using the MENT-EGAS prototype. This represents the first study of estrus detection using an EN detection by direct sampling. EN technologies, such as MENT-EGAS, could be applied in the future in dairy cattle farms as a precise, non-invasive method for estrus detection.