The amount of time cattle spend standing in streams is one input parameter needed by computer models when total maximum daily load (TMDL) plans are being developed. This input parameter is estimated using professionals' best judgment because experimental data are not available, and estimations are generally inconsistent. The goal of this study was to gain a better estimate of the amount of time cows spend in streams, since this has a significant impact on direct fecal coliform loadings to streams. Significant factors influencing the amount of time cattle spend in streams were identified, and a relationship was developed for predicting the amounts of time cattle spend in streams. Five farms were studied in southwest Virginia from August 2001 through February 2002. Camera surveillance systems were set up on two beef farms and three dairy heifer operations, and cattle activity in streams was recorded during daylight hours. Climatic data, pasture characteristics, feed characteristics, and farm management practices were collected from each site, and their relationships with the amount of time cattle spend in streams were investigated. No significant difference (p=0.82) was found between the amount of time beef cattle and dairy heifers spent in streams. Overall, cows spent an average of 10.12 min day-1cow-1 standing in streams during the observation period. Throughout the study period from August to February, cows spent the highest amount of time in streams during the month of November (14.3 min day-1cow-1). Feed, climatic, and pasture parameters were found to influence the amount of time cattle spent in streams. These significant parameters were used to develop an empirical equation for predicting cattle presence in streams. This model may not accurately predict the amount of time spent in streams by cows during warmer summer months, since data was collected during fall and winter months. Other limitations may be encountered when using the model to predict the amount of time lactating dairy cows spend in streams when streams are their sole water source, since they have a much higher water requirement than those cows studied. / Master of Science

The overall objective of this thesis was to assess the effect of feed area design including feeding space availability, barrier type and stocking density, on the feeding behaviour of dairy cows. Feed intake in dairy cows is directly related to milk production, thus a good food supply is extremely important to the modern, high producing dairy cow. Intake is critical for improving milk production, health, body condition and the welfare of the animals. Feeding designs can have a major effect on behaviour and feed intake, therefore it is an important consideration when housing cattle and other livestock. The effect of altering the amount of space allowance available at the feed-face highlighted a significant effect on feeding patterns. As the space allowance increased the number of feeding bouts also increased (P<0.001) and length of bouts decreased (P<0.05). However, when provided with extra space at the feed-face, cows did not increase their feed intake as hypothesised, possibly as a result of the differences between individual animals being masked by an overall group effect. The number of aggressive interactions decreased as the space allowance increased (P<0.001) and furthermore, the number of times individuals were displaced from the feeding area also decreased as the space allowance increased (P<0.05). Subsequently, preference tests were used as a behavioural tool to determine how individual cows perceive their feeding environment with specific emphasis on understanding what difficulties low ranking animals face at the feed-face. Subordinate cows showed a significant preference for feeding alone rather than next to a dominant when they were offered high quality feed on both sides of a Y-maze test (P<0.001). When “asked” to trade-off between feed quality and proximity to a dominant cow, subordinate cows chose to feed alone on low quality food. A follow-on experiment using the same methodology was undertaken and aimed to identify the space allowances at which cows would not trade-off food quality. Four different space allowances were tested. At the two smaller space allowances, cows preferred to feed alone and for the two larger space allowances, cows had no significant preferences. The feed barrier has been shown to have a major effect on feeding and social behaviour of group housed dairy cows. A barrier design that provides some sort of separation between cows has also been shown to reduce competition. The aim of the final study was to determine if a feed barrier which obscured the cows’ visual field whilst feeding would increase vigilance behaviour and alter normal feeding behaviour, particularly for subordinates. Two different types of feed barrier were tested at four different stocking densities. The average daily feeding time was higher when cows were fed using a conventional headlock system compared to an electronic feeding system (P<0.05). All groups of cows displayed vigilance scans, however, neither type of barrier, feed space allowance, or dominance rank had an effect on the frequency of scans. These results indicate that neither feeder design nor stocking rate affect vigilance in dairy cows, at least over the treatment conditions assessed in the current study. The results of this research illustrate that to achieve the maximum levels of feeding behaviour and a reduction of aggressive behaviour, the cows’ environment must be such that it provides sufficient space and feed barrier design which will allow normal social behaviour. Over-stocking at the feed-face should be avoided to reduce competition. Future research should consider the long term effects of over stocking and competition on dry matter intake (DMI), milk production and health.

A study was conducted in order to evaluate cattle behavior, stream bank erosion and water quality due to the installation of off-stream water sources for grazing cattle as an alternative to stream bank fencing. The study was located on two commercial cow-calf operations in southwest Virginia which utilized rotational grazing. The presence of an off-stream water source for grazing cattle greatly reduced the negative impact which grazing cattle have upon stream bank erosion and water quality. Field observations of cattle behavior indicated that cattle preferred to drink from an off-stream water source over that of an adjacent stream 92% of the time. The installation of an off-stream water source reduced the total time which cattle spend within the stream area by 58% and the amount of stream bank erosion by 76%. Due to the installation of off-stream water sources, concentrations of total suspended solids, total nitrogen, ammonium, sediment-bound nitrogen, total phosphorus and sediment-bound phosphorus, were reduced by 90%, 54%, 70%, 68%, 81% and 75%, respectively. Concentrations of fecal coliform and fecal streptococci decreased by 51% and 77% when an off-stream water source was available for grazing cattle. Lastly, an economic analysis of the data indicated that the cost of developing off-stream water sources was considerably less than that of implementing six fenced stream bank buffer zone scenarios on the two farms. / Master of Science

Edge intelligence can reduce power dissipation to enable power-hungry long-range wireless applications. This work applies edge intelligence to quantify the reduction in power dissipation. We designed a wireless sensor node with a LoRa radio and implemented a decision tree classifier, in situ, to classify behaviors of cattle. We estimate that employing edge intelligence on our wireless sensor node reduces its average power dissipation by up to a factor of 50, from 20.10 mW to 0.41 mW. We also observe that edge intelligence increases the link budget without significantly affecting average power dissipation. / Master of Science / Battery powered sensor nodes have access to a limited amount of energy. However, many applications of sensor nodes such as animal monitoring require energy intensive, long range data transmissions. In this work, we used machine learning to process motion data within our sensor node to classify cattle behaviors. We estimate that transmitting processed data dissipates up to 50 times less power when compared to transmitting raw data. Due to the properties of our transmission protocol, we also observe that transmitting processed data increases the range of transmissions without impacting power dissipation.

Small herds of cows were observed and spatially mapped over continuous twenty-four hour periods. Treatments were implemented that investigated the effect of water site and supplementation on animal distribution patterns. A series of six observation periods constituted each observation series. Observation series were repeated winter (January) and summer (July) for two years. Forage conditions varied considerably between years and seasons. Three regression models for different periods related forage variables to animal use with R�� values ranging from 0.51 to 0.77. A spatial point analysis, Ripley's K, also discerned differences in spatial point arrangements related to differences in forage and season. It detected and quantified changes caused by locating a high-protein supplement in the pasture. Swale sites and slopes of less than 10 percent were preferred for grazing in all seasons. Water sources and shade trees were distribution focal points for three observation series. In the winter of 1998, animals spent more time on warmer sites. We recorded more cow activity and movement at night than other researchers. Resting areas had aspect and elevation attributes that relate to temperature regulation. Animal positions were analyzed to determine cattle subgroups. Forage availability and thermoregulatory needs influenced the distance between associated members. Social dominance and subgroup membership were closely related to the age of individual animals. A geographic information system based technique called multi-criteria evaluation was used to develop temporal/spatial models predicting cattle distribution across the landscape. Summer models worked better than winter models because water sources and shade sites were more consistent as focal points for cattle activities. / Graduation date: 2002

Knowledge of scale-dependent factors that affect the distribution of ungulate herds, such as cattle, is essential for the development of more accurate predictive models of animal movement and the management of sustainable agriculture. Our objectives were to evaluate the seasonal and diurnal patterns of herd spread, explore grouping dynamics, and the influence of dominance ranks, seasonal changes in forage biomass, use of shade, water and supplemental feed areas on cattle distribution. Positional and activity information of eleven free-ranging Bonsmara, Bos taurus, cows were obtained at five minute intervals using Global Positioning System collars. We conducted a total of 12 trials each of three weeks from August 2007 to August 2009 in a 457ha shrub-dominated savanna in South Texas. Spread was obtained by calculating the average Euclidean distance of individuals to the center of the herd. The association software package, ASSOC1, was used to analyze spatio-temporal interactions. Normalized Difference Vegetation Index, an indicator of available forage biomass, was calculated using satellite imagery. Number of GPS fixes was used as an indicator of animal use of shade patches and water and supplemental feeding areas. In this semi-arid environment, herd spread was greatest and subgroup division occurred during summer when forage biomass was high. Herd spread was the least and shade patches were used more during winter when forage biomass was low. Throughout the year spread was smallest and use of shade patches highest during the midday period compared to other periods of the day. Location of individuals in the center of the herd was not associated with their dominance ranking. There was no significant overall pattern of seasonal changes in cattle use of water and supplemental feeding areas. Seasonal forage availability and shade patches seem to have a greater influence on herd dispersion patterns and grouping dynamics than any effect of social dominance. Herd distribution is likely related to changes in forage biomass and temperature fluctuations which drive compensatory night-time feeding and thermoregulatory actions. Accurate ethological studies of herds depend on our ability to understand herd dynamics at multiple scales that affect and are relevant to animal's response to the landscape.