|dc.description.abstract||In countries with temperate climates such as New Zealand and Ireland, grazing pasture plays an essential role in dairy production and farm profitability. In these climates, pasture growth occurs year-round in a seasonal pattern, making pasture a low-cost, high-quality feed source if well managed. However, the overall success of this system depends on on-farm management decisions, including pasture management (i.e., quality versus quantity) and meeting cow requirements (i.e., feed demand).
Knowledge of individual paddock performance and grazing information can assist with farm management decisions that maximise pasture productivity and profitability and reduce the dependence on imported feeds. Some farmers record pasture covers and grazed paddocks as part of their regular farm walks to enable feeding and farm management decisions. However, this is not common on many farms. Furthermore, these records are not necessarily kept electronically or on paper after the immediate decisions have been made. Therefore, this research aimed to determine the accuracy and precision of Global Positioning System (GPS) enabled devices (i.e., solar-powered ear tags and cow collars) and establish if they could be used to automate the recording of on-farm grazing events and determine the area allocated.
Prior to field testing, static testing of three GPS device types (Agtech and mOOvement solar-powered ear tags and Oyster2 collars) suitable for use on cattle was conducted to determine their accuracy (location error) and precision (Circle Error Probable, CEP). The mean location error (MLE) was similar for the Agtech and Oyster2 devices at 5.4 m and 5.7 m, respectively, compared with 34.2 m MLE for the mOOvement devices. In addition, the mean 95% CEP was calculated for all device types, with the Oyster2 devices having the lowest at 11.9 m, followed by Agtech at 13.9 m and mOOvement at 77.6 m.
The ability to use GPS devices to record grazing events automatically was tested on a 400-cow dairy herd in Canterbury, New Zealand. A small proportion of the milking herd was fitted with GPS enabled devices (i.e., ~2.8% Agtech & mOOvement ear tags (11 animals) and 0.5% digitanimal collars (2 animals)). According to the device fix rate, these devices recorded an animal's location at hourly or two-hourly intervals. The data was analysed for four days, from the 5ᵗʰ to the 8ᵗʰ of April 2021, for the AM (6 am – 3 pm) and PM grazing periods (3 pm – 6 am the following day). The digitanimal collars recorded 62.3% (of total observations) in the correct paddock, as opposed to adjacent paddocks or races, for the four days analysed, compared with 52.5% for the Agtech devices and 45.2% for the mOOvement ear tags.
A computer simulation of GPS fixes during a 6-hour daybreak was conducted to estimate the required number of devices and device fix rate. Four variables were adjusted across different scenarios to calculate the number of devices and the fix rate needed to identify which paddock a herd of cows were grazing correctly. The variables were a) the location error of individual GPS devices, b) paddock shape, c) paddock size, and d) the number of GPS fixes (i.e., number tags/herd * device fix rate * hours in the paddock). The proportion of GPS position fixes correctly within the paddock boundary was then calculated. The simulation, combined with the on-farm study, suggested that to identify the grazed paddock using GPS devices approximately 1% of herd size should be fitted with devices with a one hour or more frequent fix rate. For herds of less than 300 cows, it is recommended to use a minimum of three devices per herd.
The ability of GPS devices to determine the grazing area allocated was tested using a further simulation exercise. This was done by dividing a simulated one-hectare square paddock into sixteen equal-sized squares. The method showed promise but was sensitive to the fix threshold selected (i.e., number of fixes per square). In general, a higher fix threshold is required to estimate the area allocated accurately with reduced area allocated. Therefore, if attempting to estimate the area allocated, a higher fix rate is needed than for paddock identification.
This preliminary investigation has shown that it is feasible to use GPS devices to determine the area allocated and record grazing events on-farm depending on the device type and settings. For example, to identify the grazed paddock, it is appropriate for ~1% of the herd size to be fitted with devices; however, to determine the area allocated, a higher number of devices is likely required ~3% of herd size. Nevertheless, before this technology could become standard on-farm practice, further work is required, including a comprehensive on-farm study. This is needed to provide better data to assess the accuracy and precision of the individual devices and other potential on-farm complications, including paddock shape, use of crops, and multiple herds during the winter and early spring periods (e.g., dry cows, springers, colostrums and milkers). Nonetheless, with a greater focus on maximising pasture utilisation on dairy farms today and the increasing range (e.g., direct to satellite options) and decreasing cost of GPS devices, there is likely to be greater interest in this technology in the future.||