Using smart farming technologies to improve the sustainabilty of livestock grazing systems
Kumulative Dissertation
Datum der mündl. Prüfung:2022-12-16
Erschienen:2023-01-12
Betreuer:Dr. Martin Komainda
Gutachter:Prof. Dr. Johannes Isselstein
Gutachter:Prof. Dr. Imke Traulsen
Gutachter:Prof. Dr. Georg Bareth
Dateien
Name:Dissertation.pdf
Size:44.6Mb
Format:PDF
Zusammenfassung
Englisch
Semi-natural grasslands, which contribute to a wide range of ecosystem services, such as soil erosion control, water storage and carbon storage, are severely threatened by both intensification and complete extensification, i.e. the abandonment of agricultural management. The genesis of these open landscapes is inextricably linked to grazing livestock, which in Europe was mainly herded on common pastureland by herders until the 19th century. The immense demand for labour for this form of pastoralism was met, among other things, at ‘Hütekindermärkte’. With the simplification of fencing due to the invention of barbed wire in 1873 and electric fencing in 1937, the loss of common pastureland, the improved possibilities for preserving forage and the increasing lack of herdsmen, more and more pastureland was fenced. However, the main task of the herder, the continuous monitoring of the grazing livestock and the sward could not be taken over by fences. For successful grazing, the optimal spatial and temporal allocation of the grazing animal to the sward is crucial. The grazing animal influences the sward just as the sward influences the grazing animal. The lack of consideration of this relationship in grazing management leads to semi-natural grasslands being damaged by overgrazing, especially in arid areas, while on the other hand the abandonment of grazing (mainly in Europe) leads to natural succession and thus to scrub encroachment of the formerly open landscape. Since the invention of the electric fence, there have been no major developments in the field of pasture management that could have led to improved spatial and temporal allocation of grazing animals on the sward. The development of GPS-based virtual fencing, where each adult animal wears a special GPS collar, and grazing allocation takes place via a mobile device, seems to offer great possibilities for a more precise and flexible grazing system. Fencing could evolve again more towards holistic herding. In this context, the aim of this work is to generate a knowledge gain in the existing system, as well as a test and evaluation of options for an innovative herding fencing system by using virtual fences. A kind of herdsman 4.0, a system which should be able to take into account the needs of the animals and the sward, as well as to enable a spatially and temporally continuous monitoring of the animals. In order to analyse possible correlations between forage availability and movement behaviour and thus pasture use, chapter one of this thesis uses the long-term grazing trial 'FORBIOBEN' with its three different grazing intensities to investigate hourly and daily walking distances as well as the spatial distribution of Fleckvieh cows by the use of commonly used GPS collars for animal monitoring. The activity of the cows increased with lower forage availability and the spatial distribution during the active time of the animals, identified by daily hydrographs, is more even at the highest grazing intensity. It should be noted, however, that in this study the animals increased their activity in both the highest and the lowest grazing intensity. There are indications that heterogeneity and, generally speaking, the distribution of forage resources on the pasture must be taken into account in order to be able to make reliable statements about animal movement as a reflection of forage availability on the pasture. Chapter two addresses the use of virtual fences for the first time and investigates whether there is a negative effect of virtual fencing technology on Fleckvieh heifers. Continuous animal observations, faecal samples, grassland measurements and steps walked provided no evidence of deterioration compared to the control group. Furthermore, it could be shown that despite individual differences, all animals were able to adjust to the virtual fencing system. The exclusion of the virtually fenced animals was successful. No animal crossed the virtual boundary during the experiment. Animal monitoring in the pasture may be a possible additional benefit of the GPS collars used for virtual fencing. Chapter three investigates in a model way correlations between spatiotemporally explicit animal behaviour and data from drone-based remote sensing of the grazing area in the context of a rotational grazing trial. This information, which has been validated as part of the study, may in combination provide information on hotspots of cattle residence time, biomass changes and animal behaviour. The comparison of the lying times identified by the collars with animal observation data provides accurate results. The relationship between RGBVI from UAV and biomass was significant with a moderate amount of explained variability. Overall, it appears that animal monitoring on pasture is possible through virtual fence collars and can provide highly valuable data presenting the GPS location of behavioural events. Distances walked by animals, their distribution over the area and lying times per day show an effect of decreasing forage availability. In this context, lying times decreased, walking distances increased and the distribution of animals on the area became more even. The drone data of the grazing areas before and after grazing showed a clear correlation to the local stays of the animals associated with grazing on the studied, predefined 2.5 × 2.5 m polygon grid. Further research is needed to be able to use the whole relationship between animal behaviour on pasture and drone data to define, among other things, thresholds for forage scarcity. Overall, the results of the individual chapters build on each other and could provide basic requirements for the development of an innovative virtual herding system. An effect of forage availability on the movement behaviour of the animals could be demonstrated. The welfare of the virtually fenced animals, compared to the conventionally fenced animals is not affected in our trial, although further research over longer periods is needed to further validate the results. Drone-based remote sensing to record an actual condition of the pastures can provide a baseline that could lead to better detection of available biomass through continuous monitoring of animal residency. An appropriately adapted virtual herding system (Herder 4.0) can help to simplify and improve the diverse challenges of pasture management in a sustainable way. An example of this is soil degradation. With the help of the animal residence times from the collar data, which are verified with the UAV images and give an indication of the change in the sward, it may be possible to stop such processes or to react quickly with virtual out fencing of the heavily used areas. This high flexibility allows to rethink the previously known grassland systems, which are strongly oriented towards the possibilities of fencing. The technology of virtual fencing in combination with remote sensing enables fine-scale ‘grid grazing’ that effectively and sustainably considers grazing animals and swards.
Keywords: Virtual fencing; Precision livestock farming; Pasture management; Animal welfare