Feed Intake

The experimental design was evaluated by the Animal Welfare Body of the University of Bologna and found as not falling within the Directive 63/2010 of the European Parliament and of the Council on the protection of animals used for scientific purposes (transposed into Italian law by Legislative Decree 26/2014), thus not requiring any authorization from the national competent Authorities. The study was approved by the Animal Ethics Committee of the University of Bologna and conducted at the dairy cattle farm of the same university. Animals were 8 lactating Holstein–Friesian cows (average weight 752 ± 52 kg) in their late lactation period (262 ± 49 days in milk), with an average milk yield of 19.6 ± 6.6 kg/day. Cows were housed in a naturally ventilated free stall barn and had free access to feed and water. Rations were formulated to mimic total mixed rations used in the Parmigiano Reggiano cheese production area, in Italy, which consisted of all dry and nonfermented components, as detailed in previous studies [50 (link),51 (link)].
Animals were divided into two groups, consisting of 4 cows each, and two different treatments of 10 days’ duration were performed. Group AFB1 received, from day 0 to day 2 only, the basal diet (AFB1 content less than 2 ng/g, see Section 4.3), while group TP was offered the basal diet supplemented with 20 g turmeric powder dissolved in linseed oil/head/day for 10 consecutive days. At day 3, all cows received the basal diet containing naturally contaminated maize with a final AFB1 concentration of 5 ± 1 μg/kg for 8 consecutive days. A crossover experimental design was applied: each cow received both treatments sequentially after of a 4-day washout period, during which all animals were offered the basal diet. The used TP contained 2.5% curcumin, desmethoxycurcumin and bis-desmethoxycurcumin (85/10/5). With an average feed intake of 20 kg/cow/day, the cows approximately received 0.5 g active substances/head/day; this dosage is higher than the recommended inclusion level for flavoring purposes but below the maximum safe concentration of 0.72 g/day calculated by the EFSA [7 (link)].
Animals were milked twice a day, namely at 08:00 h (referred to as M) and 19:30 h (referred to as E) in a double-5 herringbone milking parlour. Individual milk samples (around 100 mL) collected at T0, T2, T4, T6, T8 and T10 were used for the determination of somatic cell count (SCC) (Fossomatic 7, Foss Electric A/S, Hillerød, Denmark) and milk composition (fat, lactose, protein, urea) by means of infrared spectroscopy (MilkoScan FT+, Foss Electric A/S, Hillerød, Denmark). The determination of AFB1 metabolites was performed on further milk samples (around 100 mL) collected at T0 (M), T4, T5 and T6 (M and E), and T7, T8, T9 and T10 (M only), which were stored frozen (−20 °C) pending analysis. The experimental design is outlined in Figure 2.

Sixty crossbred yearling Angus heifers (initial BW = 400 ± 6 kg) were managed as a single pasture group with free access to graze native range and were randomly assigned to 1 of 3 dietary treatments 1) no access to feed supplements (CON; N = 20), 2) free choice access to mineral supplement (MIN; Purina Wind & Rain Storm All-Season 7.5 Complete, Land O’Lakes, Inc., Arden Hills, MN, N = 20), or 3) free choice access to energy supplement (NRG; Purina Accuration Range Supplement 33, Land O’Lakes, Inc., Arden Hills, MN, N = 20). The manufacturer recommendation for daily intake of the mineral supplement was 113 g. The NRG supplement was formulated by adding 68.1 kg MIN to a 907.4 kg mixture of 60% ground corn and 40% Accuration (25.5 % CP; Table 1) with an anticipated daily intake of 1.63 kg. Thus, if heifers in the NRG treatment consumed 1.63 kg of supplement, and heifers in the MIN treatment consumed 113 g, then both the MIN and NRG heifers would be consuming the same amount of the mineral product used. The MIN and NRG supplements were delivered via the MCCC units which were located within 50 m of the waterer in the pasture. Feeders were set to restrict access of CON heifers from either trailer unit, with MIN and NRG heifers having ad libitum access to the trailer containing their respective feed assignment. Because few heifers consumed either supplement early in the grazing season (Figure 1), feed intake data were summarized over a 57-d period; from the time of pregnancy diagnosis (July 25) until removal from pasture (September 19).
The CowManager system reported the minutes spent during each hour of every day in activity categories including “eating”, “ruminating”, “not active”, “active”, and “highly active”, with a proprietary model and available through the web-based application. Estrus-related alerts were continuously generated via the CowManager system, including classifications of “in heat”, “potential”, or “suspicious”. Pregnancy detection was performed 34 d after AI via transrectal ultrasonography (7.0-MHz transducer, 500 V Aloka, Wallingford, CT). Continuous monitoring with the CowManager tag provided data related to heifer estrus activity. A retrospective analysis was conducted to determine the accuracy of estrus-related alerts generated via the CowManager system versus a known pregnancy status determined via ultrasound. Similarly, a retrospective analysis was conducted to evaluate the accuracy of health events that were flagged via the CowManager system (reported as “sick”, “very sick”, or “no movement”) by comparing electronic alerts with treatment logs generated by the animal care staff. It is important to note that the CowManager system has been validated using the proprietary algorithm in populations of dairy cows housed indoors (Bikker et al., 2014 (link)) and grazing (Pereira et al., 2018 (link)).

Each of two Mobile Cow Command Center (MCCC) units were developed by pairing two commercially available technologies into single trailer units that can be transported and function anywhere cattle are managed. The first technology is the SmartFeed device (C-lock Inc., Rapid City, SD), which is a self-contained system designed to measure supplement intake and feeding behavior from individual cattle in group settings. The system is solar powered and includes a radio-frequency identification (RFID) reader, weigh scales, access control gate, a feed bin, and a cloud-based interface which continuously logs feed intake and feeding behavior data. The programming of the SmartFeed units is flexible, with the ability to assign specific animals to specific feeders and to prohibit entry of individual animals once a daily target intake is achieved. The second technology included in the MCCC was the CowManager system (CowManager B.V., the Netherlands), which fits over an RFID ear tag and uses additional sensors to monitor cow reproductive (estrus alerts), feeding-related (eating, rumination, and activity level), and health-associated data. The CowManager ear tag continuously registers movements from the cow’s ear and classifies the data through proprietary algorithms (Pereira et al., 2018 (link)). Data are sent through a wireless connection, via a router placed on the top of the MCCC unit. Data are then received through a coordinator unit that is attached to a computer in a lab (approximately 200 m line of site from the MCCC units) that automatically uploaded the data for viewing on any device with an internet connection. Each MCCC contained 2 SmartFeed units, controlling hardware and the CowManager router in an enclosed trailer with open feed access areas and retractable wheels for transport.