Automatic Lameness Detection in a Milking Robot : Instrumentation, measurement software, algorithms for data analysis and a neural network model

The aim of this thesis is to develop a fully automatic lameness detection system that operates in a milking robot. The instrumentation, measurement software, algorithms for data analysis and a neural network model for lameness detection were developed. Automatic milking has become a common practice in dairy husbandry, and in the year 2006 about 4000 farms worldwide used over 6000 milking robots. There is a worldwide movement with the objective of fully automating every process from feeding to milking. Increase in automation is a consequence of increasing farm sizes, the demand for more efficient production and the growth of labour costs. As the level of automation increases, the time that the cattle keeper uses for monitoring animals often decreases. This has created a need for systems for automatically monitoring the health of farm animals. The popularity of milking robots also offers a new and unique possibility to monitor animals in a single confined space up to four times daily. Lameness is a crucial welfare issue in the modern dairy industry. Limb disorders cause serious welfare, health and economic problems especially in loose housing of cattle. Lameness causes losses in milk production and leads to early culling of animals. These costs could be reduced with early identification and treatment. At present, only a few methods for automatically detecting lameness have been developed, and the most common methods used for lameness detection and assessment are various visual locomotion scoring systems. The problem with locomotion scoring is that it needs experience to be conducted properly, it is labour intensive as an on-farm method and the results are subjective. A four balance system for measuring the leg load distribution of dairy cows during milking in order to detect lameness was developed and set up in the University of Helsinki Research farm Suitia. The leg weights of 73 cows were successfully recorded during almost 10,000 robotic milkings over a period of 5 months. The cows were locomotion scored weekly, and the lame cows were inspected clinically for hoof lesions. Unsuccessful measurements, caused by cows standing outside the balances, were removed from the data with a special algorithm, and the mean leg loads and the number of kicks during milking was calculated. In order to develop an expert system to automatically detect lameness cases, a model was needed. A probabilistic neural network (PNN) classifier model was chosen for the task. The data was divided in two parts and 5,074 measurements from 37 cows were used to train the model. The operation of the model was evaluated for its ability to detect lameness in the validating dataset, which had 4,868 measurements from 36 cows. The model was able to classify 96% of the measurements correctly as sound or lame cows, and 100% of the lameness cases in the validation data were identified. The number of measurements causing false alarms was 1.1%. The developed model has the potential to be used for on-farm decision support and can be used in a real-time lameness monitoring system.

Interactions between transgenic trees and mycorrhizal and pathogenic fungi

The development of biotechnology techniques in plant breeding and the new commercial applications have raised public and scientific concerns about the safety of genetically modified (GM) crops and trees. To find out the feasibility of these new technologies in the breeding of commercially important Finnish hardwood species and to estimate the ecological risks of the produced transgenic plants, the experiments of this study have been conducted as a part of a larger project focusing on the risk assessment of GM-trees. Transgenic Betula pendula and Populus trees were produced via Agrobacterium mediated transformation. Stilbene synthase (STS) gene from pine (Pinus sylvestris) and chitinase gene from sugar beet (Beta vulgaris) were transferred to (hybrid) aspen and birch, respectively, to improve disease resistance against fungal pathogens. To modify lignin biosynthesis, a 4-coumarate:coenzyme A ligase (4CL) gene fragment in antisense orientation was introduced into two birch clones. In in vitro test, one transgenic aspen line expressing pine STS gene showed increased resistance to decay fungus Phellinus tremulae. In the field, chitinase transgenic birch lines were more susceptible to leaf spot (Pyrenopeziza betulicola) than the non-transgenic control clone while the resistance against birch rust (Melampsoridium betulinum) was improved. No changes in the content or composition of lignin were detected in the 4CL antisense birch lines. In order to evaluate the ecological effects of the produced GM trees on non-target organisms, an in vitro mycorrhiza experiment with Paxillus involutus and a decomposition experiment in the field were performed. The expression of a transgenic chitinase did not disturb the establishment of mycorrhizal symbiosis between birch and P. involutus in vitro. 4CL antisense transformed birch lines showed retarded root growth but were able to form normal ectomycorrhizal associations with the mycorrhizal fungus in vitro. 4CL lines also showed normal litter decomposition. Unexpected growth reductions resulting from the gene transformation were observed in chitinase transgenic and 4CL antisense birch lines. These results indicate that genetic engineering can provide a tool in increasing disease resistance in Finnish tree species. More extensive data with several ectomycorrhizal species is needed to evaluate the consequences of transgene expression on beneficial plant-fungus symbioses. The potential pleiotropic effects of the transgene should also be taken into account when considering the safety of transgenic trees.

Concentrate feeding strategies for growing and finishing dairy bulls offered grass silage-based diets

The first aim of this thesis was to produce data for evaluating, developing and recommending biologically and economically efficient energy and protein feeding strategies for growing and finishing dairy bulls offered grass silage-based diets. The second aim was to calculate the energy and protein supplies of the dairy bulls fed different grass silage-cereal-based diets and, based on this, to estimate the possible need to revise the current Finnish energy and protein recommendations for growing dairy bulls. The third aim was to demonstrate the phosphorus supply of dairy bulls fed grass silage-cereal-based diets with or without protein supplementation in relation to current feeding recommendations for phosphorus. The results indicate that protein supplement is not needed for finishing dairy bulls (live weight more than 250 kg) when they are fed good-quality grass silage (digestible organic matter more than 650 g/kg dry matter, restricted fermentation with low concentrations of fermentation acids and ammonia N) and grain-based concentrate with a moderate (300-700 g/kg dry matter) concentrate level. The results also suggest that with total mixed ration feeding it is possible to use rather high concentrate proportions (700 g/kg dry matter) in feeding dairy bulls. According to this study, barley fibre is a suitable energy supplement with good-quality silage for growing dairy bulls. The results suggest that 50% of barley grain can be replaced with barley fibre without affecting growth. Also oats is a suitable energy supplement for dairy bulls. However, as a consequence of decreased energy intake, the gain and feed conversion of the bulls were slightly reduced in this study when barley grain was replaced by oats in the diet. Ultimately, the rationality of the use of barley fibre and oats in the future will depend on the price in relation to other concentrates. During the feeding experiments the calculated supply of energy was 10% higher than in the Finnish feeding recommendations for the present growth rate. This indicates that there is a need to update the Finnish feeding recommendations for dairy-breed growing bulls, and further calculations are needed for the energy supply of growing dairy bulls. The calculated supply of AAT (amino acids absorbed from the small intestine) was 38% higher than in the Finnish feeding recommendations for the present growth. Possibly, the present AAT-PBV system is not an optimal protein evaluation system for growing dairy bulls more than 250 kg live weight. The calculations based on the feeding experiments and the Finnish feeding recommendations indicate that in most cases the dairy bulls (live weight more than 250 kg) received enough P from the basic grass silage cereal-based diets without additional mineral feeds. Therefore there is no need to add P in the form of mineral mixtures.

Sweet potato chlorotic stunt virus: Studies on viral synergism and suppression of RNA silencing

The studies presented in this thesis aimed to a better understanding of the molecular biology of Sweet potato chlorotic stunt virus (SPCSV, Crinivirus, Closteroviridae) and its role in the development of synergistic viral diseases. The emphasis was on the severe sweet potato virus disease (SPVD) that results from a synergistic interaction of SPCSV and Sweet potato feathery mottle virus (SPFMV, Potyvirus, Potyviridae). SPVD is the most important disease affecting sweetpotato. It is manifested as a significant increase in symptom severity and SPFMV titres. This is accompanied by a dramatic sweetpotato yield reduction. SPCSV titres remain little affected in the diseased plants. Viral synergistic interactions have been associated with the suppression of an adaptive general defence mechanism discovered in plants and known as RNA silencing. In the studies of this thesis two novel proteins (RNase3 and p22) identified in the genome of a Ugandan SPCSV isolate were shown to be involved in suppression of RNA silencing. RNase3 displayed a dsRNA-specific endonuclease activity that enhanced the RNA-silencing suppression activity of p22. Comparative analyses of criniviral genomes revealed variability in the gene content at the 3´end of the genomic RNA1. Molecular analyses of different isolates of SPCSV indicated a marked intraspecific heterogeneity in this region where the p22 and RNase3 genes are located. Isolates of the East African strain of SPCSV from Tanzania and Peru and an isolate from Israel were missing a 767-nt fragment that included the p22 gene. However, regardless of the absence of p22, all SPCSV isolates acted synergistically with SPFMV in co-infected sweetpotato, enhanced SPFMV titres and caused SPVD. These results showed that p22 is dispensable for development of SPVD. The role of RNase3 in SPVD was then studied by generating transgenic plants expressing the RNase3 protein. These plants had increased titres of SPFMV (ca. 600-fold higher in comparison with nontransgenic plants) 2-3 weeks after graft inoculation and displayed the characteristic SPVD symptoms. RNA silencing suppression (RSS) activity of RNase3 was detected in agroinfiltrated leaves of Nicotiana bethamiana. In vitro studies showed that RNase3 was able to cleave small interferring RNAs (siRNA) to products of ~14-nt. The data thus identified RNase3 as a suppressor of RNA silencing able to cleave siRNAs. RNase3 expression alone was sufficient for breaking down resistance to SPFMV in sweetpotato and for the development of SPVD. Similar RNase III-like genes exist in animal viruses which points out a novel and possibly more general mechanism of RSS by viruses. A reproducible method of sweetpotato transformation was used to target RNA silencing against the SPCSV polymerase region (RdRp) with an intron-spliced hairpin construct. Hence, engineered resistance to SPCSV was obtained. Ten out of 20 transgenic events challenged with SPCSV alone showed significantly reduced virus titres. This was however not sufficient to prevent SPVD upon coinfection with SPFMV. Immunity to SPCSV seems to be required to control SPVD and targeting of different SPCSV regions need to be assessed in further studies. Based on the identified key role of RNase3 in SPVD the possibility to design constructs that target this gene might prove more efficient in future studies.

The virulence of Finnish Pyrenophora teres f. teres isolates and its implications for resistance breeding

In Finland, barley, Hordeum vulgare L., covers 50 % of the total acreage devoted to cereal cultivation. The most common disease of barley in Finland is net blotch, a foliar disease caused by the ascomycete Pyrenophora teres Drechsler. Disease resistance based on plant genes is an environmentally friendly and economical way to manage plant diseases caused by biotic stresses. Development of a disease resistance breeding programme is dependent on knowledge of the pathogen. In addition to information on the epidemiology and virulence of a pathogen, knowledge on how the pathogen evolves and the nature of the risks that might arise in the future are essential issues that need to be taken into account to achieve the final breeding aims. The main objectives of this study were to establish reliable and efficient testing methods for Pyrenophora teres f. teres virulence screening, and to understand the role of virulence of P. teres f. teres in Finland from a disease resistance breeding point of view. The virulence of P. teres was studied by testing 239 Finnish P. teres f. teres isolates collected between 1994 2007 originating from 19 locations, and 200 P. teres progeny isolates originating from artificially produced P. teres matings. According to the results of this study, screening for P. teres f. teres isolates on barley seedlings under greenhouse conditions is a feasible and cost efficient method to describe the virulence spectrum of the pathogen. Inoculum concentration and the seedling leaf used to gauge virulence had significant effects. Barley grain size, morphological traits of P. teres isolates, spore production and growth rate on agar did not affect the expression of virulence. A common barley differential set to characterize the P. teres virulence was developed and is recommended to be used globally. The virulence spectrum of Finnish P. teres f. teres isolates collected in 1994-2007 was constant both within and between the years. The results indicated differences in the pathogen s aggressiveness and in barley genotypes resistance. However, differences in virulence were rarely significant. Unlike in laboratory conditions, no indications of changes in virulence caused by the sexual reproduction have been observed in Finnish barley fields. In Finland, durable net blotch resistance has been achieved by introducing resistance from other barley varieties using traditional crossing methods, including wide crossing, and testing the breeding material at early generations at several sites under natural infection pressure. Novel resistance is available, which is recommended to minimize the risk of selection of virulent isolates and breakdown of currently deployed resistance.

Effects of seed origin latitude on the timing of height growth cessation and field performance of silver birch

The aim of this thesis was to increase our knowledge about the effects of seed origin on the timing of height growth cessation and field performance of silver birch from different latitudes, with special attention paid to the browsing damage by moose in young birch plantations. The effect of seed origin latitude and sowing time on timing of height growth cessation of first-year seedlings was studied in a greenhouse experiment with seven seed origins (lat. 58º - 67ºN). Variation in critical night length (CNL) for 50 % bud set within two latitudinally distant stands (60º and 67ºN) was studied in three phytotron experiments. Browsing by moose on 5-11 -year-old silver birch saplings from latitudinally different seed origins (53º - 67ºN) was studied in a field experiment in southern Finland. Yield and stem quality of 22-year-old silver birch trees of Baltic, Finnish and Russian origin (54º - 63ºN) and the effect of latitudinal seed transfers were studied in two provenance trials at Tuusula, southern and Viitasaari, central Finland. The timing of height growth cessation depended systematically on latitude of seed origin and sowing date. The more northern the seed origin, the earlier the growth cessation and the shorter the growth period. Later sowing dates delayed growth cessation but also shortened the growth period. The mean CNL of the southern ecotype was longer, 6.3 ± 0.2 h (95 % confidence interval), than that of the northern ecotype, 3.1 ± 0.3 h. Within-ecotype variance of the CNL was higher in the northern ecotype (0.484 h2) than in the southern ecotype (0.150 h2). Browsing by moose decreased with increasing latitude of seed origin and sapling height. Origins transferred from more southern latitudes were more heavily browsed than the more northern native ones. Southern Finnish seed origins produced the highest volume per unit area in central Finland (lat. 63º11'N). Estonian and north Latvian stand seed origins, and the southern Finnish plus tree origins, were the most productive ones in southern Finland (lat. 60º21'N). Latitudinal seed transfer distance had a significant effect on survival, stem volume/ha and proportion of trees with a stem defect. The relationship of both survival and stem volume/ha to the latitudinal seed transfer distance was curvilinear. Volume was increased by transferring seed from ca. 2 degrees of latitude from the south. A longer transfer from the south, and transfer from the north, decreased the yield. The proportion of trees with a stem defect increased linearly in relation to the latitudinal seed transfer distance from the south.