1,720,979 research outputs found
Evaluating digital tools for welfare assessments of Atlantic salmon
No full textFish welfare is a critical issue in the farmed salmon industry. During the farming pro- cess, a significant number of fish suffer from winter sores and stress, resulting in wounds on their bodies. In this study, hyperspectral imaging (HSI), which is an emerging tech- nology, was used for wound identification. The spectral unmixing model was applied to total hemoglobin (TH) and different thresholds of TH were used to identify the wounds. The results were then compared to the YOLOv5 machine learning algorithm, which was considered the gold standard. The results showed that a TH greater than 0.6 can be used as a threshold that is almost comparable to YOLOv5 in identifying major wounds, while also effectively detecting minor wounds. In addition, a TH between 0.5 and 0.6 was identified as a potential threshold for detecting some minor wounds not detected by YOLOv5. These were potentially minor stress-induced wounds compared to major winter sore-induced wounds. In addition, statistical analysis of the potential effect of some factors, such as sex, showed that females tended to have larger wounds than males when TH was greater than 0.6. Conversely, the trend was reversed when TH was be- tween 0.5 and 0.6, and males had larger wounds than females. In the case of the light regimes, the fish exposed to 8 hours of light and 16 hours of darkness tended to have the smallest wound sizes overall. There was no obvious pattern or correlation between body weight and wound size. Similarly, there was no apparent correlation between condition factor and wound size.
HSI proved to be more accurate in detecting wounds due to its spectral information compared to the traditional method of human evaluation and even machine learning al- gorithms such as YOLOv5 on RGB images. However, further research is needed to improve its applicability in aquacultureM-D
Critical assessment of the “internal reference” method to eliminate non-genetic effects within a Combined Family Selection program on the abalone species (Haliotis midae)
Full text linkThesis (MScAgric)--Stellenbosch University, 2013.ENGLISH ABSTRACT: The aim of this study was to critically assess the implementation of the internal reference
method within the most recent 173 full-sibling growth trial of the Innovation Fund Abalone
Breeding Project. The trial was conducted over two locations for a period of five years, with
minimal replication for the majority of test families and a single full-sibling family was entered
into each experimental unit (basket) as an internal reference group. The primary focus was
firstly, to validate the performance of the internal reference group as a control for
comparisons and correction of environmental variation in test family performances.
Secondly, to identify areas of weakness and either make recommendations to remedy areas
of weakness or justify devoting resources to alternative methods of reducing extraneous
environmental variance with limitations on replication. The efficiency and statistical power
associated with utilising internal reference information as a covariate and for manual
correction respectively were examined for the 6 full-sibling test families that were replicated.
This study reports on the evaluation of factors which are potential sources of bias in
the internal reference method, the first of which, tag loss, was found to be significant after 6-
12 months. However, it was not found to bias internal reference group performances as the
factors which contribute to tag loss were found to act randomly. Variability in size ratio of
internal reference to test family at co-stocking proved a significant source of bias, as
reference groups smaller than their test family counterparts had reduced performances.
Testing for genotype by environment interactions was precluded due to the inherent lack of
replication and the subsequent confounding of genotype effects with inter-rearing structure
effects at one of the locations. However, significant differences were detected for both traits
of interest of the internal reference group over the two locations. Significant antagonistic
interactions were detected and identified as a source of bias for average daily weight gain of
replicate test families. The evaluation of average daily length gain for the efficiency of adjustment when the
internal reference is a covariate and the change in statistical power when the internal
reference is used for a manual correction, yielded conflicting results. The latter shows a
decrease in statistical power and the former shows an increase in efficiency, both resulting in
poor goodness of fit in the respective models. There was however evidence that when no
antagonistic interactions occurred “between replicate variance” decreased and therefore the
internal reference method has statistical merit provided all critical success factors are
satisfied.
Recommendations were made for future implementation of the internal reference
method to facilitate adequate statistical testing for sources of bias and the prevention
thereof. Additionally, an alternative method which may have merit in decreasing
environmental variance and the need for replication, is discussed.AFRIKAANSE OPSOMMING: Die doel van die studie was om die gebruik van ʼn interne verwysingsgroep te ontleed, soos
toegepas tydens die evaluering van 173 volsib families as deel van die Innovasiefonds
Perlemoen Teelprogram. Die evaluering is gedoen op twee lokaliteite oor 'n tydperk van vyf
jaar, met minimale replikasie van die toets families en die gebruik van ‘n enkele volsib familie
as 'n interne verwysingsgroep in elke eksperimentele eenheid (mandjie). Die primêre fokus
was eerstens om die gebruik van die interne verwysingsgroep vir die korreksie van
omgewingsvariasie in die toets familie optredes te evalueer. Tweedens, om spesifieke
gebreke te identifiseer ten opsigte van die gebruik van die interne verwysingsgroep en
aanbevelings maak dit reg te stel en om die meriete van alternatiewe metodes te oorweeg.
Die doeltreffendheid en statistiese onderskeidingsvermoë van die gebruik van interne
verwysingsgroep as 'n kovariaat is ondersoek met betrekking tot die 6 volsib groepe wat oor
voldoende replikasies beskik het.
Die studie doen voorts verslag oor die evaluering van potensiële oorsake van
sydigheid ten opsigte van die gebruik van die interne verwysingsgroep, insluitend die
beduidende verlies van identifikasie vanaf 6 tot 12 maande. Geen aanduiding van sydigheid
is egter gevind en die aanleidende oorsake van verlies van identifikasie blyk van ʼn
ewekansige aard te wees. Verskille in die grootte tussen die interne verwysingsgroep en
toets-families met aanvang van evaluering blyk 'n belangrike bron van sydigheid te wees,
waar die kleiner groepering aan verminderde prestasie gekoppel word. Bepaling van
genotipe-omgewing-interaksies kon nie uitgevoer word nie as gevolg van die inherente
gebrek van replisering oor lokaliteite. Beduidende verskille is egter waargeneem tussen
interne verwysingsgroepe oor die twee lokasies ten opsigte van die beide groei eienskappe.
Beduidende antagonistiese interaksies is waargeneem en geïdentifiseer as 'n bron van
sydigheid ten opsigte van die gemiddelde daaglikse gewigstoename van replikaat
toetsfamilies.
Die evaluering van gemiddelde daaglikse lengtetoename met die interne
verwysingsgroep as is 'n kovariaat en die verandering in statistiese ontledingsvermoë tydens
die gebruik van die interne verwysingsgroep het teenstrydige resultate opgelewer.
Laasgenoemde toon 'n afname in statistiese ontledingsvermoë en die eersgenoemde toon 'n
toename in doeltreffendheid, met beide swak passing op die onderskeie modelle. In die
afwesigheid van antagonistiese interaksies tussen replikasies het variansie afgeneem en
beskik die interne verwysingsgroep oor die nodige statistiese meriete indien daar aan al die
kritiese vereistes voldoen word. Aanbevelings is gemaak ten opsigte van die toekomstige implementering van die
interne verwysingsmetode met verwysing na voldoende statistiese toetsing vir bronne van
sydigheid en die voorkoming daarvan. 'n Verdere metode wat oor die nodige meriete beskik
om die omgewingsvariasie en die noodsaaklikheid vir replikasie te verminder, word
bespreek.Master
Genetic control of methane emission, feed efficiency and metagenomics in dairy cattle
No full textThe dairy industry faces the challenges of increasing production, remaining economically viable whilst simultaneously minimising impacts on the environment. The cost of feed is the highest variable cost of milk production, thus, improved feed efficiency is a strong wish. However, CH4 is a potent greenhouse gas with an energy value estimated as 2 -12% of the gross feed energy intake and thus represents a loss. There is, therefore, a need to identify the phenotypic and genetic relationships between efficiency of feed utilisation and CH4 production to ensure optimal breeding methods of increasing profitability and limiting environmental impact of dairy production. Feed is degraded and CH4 is produced by rumen microbes and not by the cow. The mechanisms which influence the composition of the rumen microbial community and how they, in turn, influence the feed efficiency and CH4 production of the host, are not well understood. Among the possible strategies, selective breeding has the benefit over others by being cumulative and persistent over generations. Genetic improvement through selection requires that phenotypes are recorded on large numbers of animals. Moreover, phenotypes must show variation, a portion of which must be genetic, and must have economic or societal value. Understanding the genetic co-variation behind and between these measures is crucial to simultaneous breeding for a more profitable and climate friendly dairy industry. However, the measurement of CH4 emissions, feed efficiency and the rumen microbiome under commercial conditions on a large scale is not a trivial task. The aim of this PhD project was to develop and integrate phenotyping measures for CH4 emission, feed efficiency and the rumen microbiome and to investigate their genetic potential for selective breeding. Firstly, in Chapter 2, improvements where made to the sniffer method of CH4 breath concentration recording in dairy cattle during automatic milking. An algorithm was developed to efficiently detect and correct for variable and random drift in time series between instruments and to detect when the cow’s head is out of the feed bin. Using linear mixed model methodology, repeated measures per cow were used to improve precision and control sources of inaccuracy such as sensor drift, background gas concentrations and diurnal variation, that were subsequently removed. Resultantly, highly repeatable phenotypes where obtained which demonstrated adequate agreement for the interchangeable use of two instruments. In Chapter 3, the ranking of cows under commercial conditions using the sniffer method was compared with the “gold standard” respiration chambers. Individual level correlations estimated as proxies for genetic correlations revealed a high correlation between sniffer-predicted CH4 production and CH4 production in the RC. These findings offer a proof of concept that sniffer CH4 phenotypes recorded over a week of lactation show substantial promise as large scale indicator traits for CH4 production using RC. In Chapter 4, genetic parameters were estimated between feed intake, milk production and CH4 breath concentration from sniffers over the course of the first lactation in Holstein cows in Denmark and The Netherlands. Through combining data between countries, genetic residual feed intake and breath gas concentrations were found to be significantly heritable, demonstrating that genetic improvement of feed efficiency and CH4 breath gas concentration is feasible in dairy cattle. The estimated genetic correlations from the largest dataset indicated that improved feed efficiency will also result in decreased gas emissions. Furthermore, including the breath gas concentrations in a multitrait genetic evaluation increased the accuracy of bull breeding values for gRFI, demonstrating an indirect economic value of CH4 and CO2 breath concentration phenotypes. In Chapter 5, we estimated the relative abundance of rumen bacteria and archaea and found a portion of these to be heritable in dairy cattle. The results demonstrate that host additive genetics has an influence on the abundance of some rumen bacteria and archaea. We detected significant associations between certain bacterial genera and differences in CH4 production of the host cow, further contributing to knowledge of the underlying biological mechanisms driving CH4 production of the host. We further extended quantitative genetic methods to estimate rumen microbial kinships between cows in place of additive genetic relationships. This enabled the quantification of variation in host CH4 production explained by the rumen microbial composition, expressed in the new term ‘microbiability’, as the relative proportion of host variation explained by associated microbes. Crucially the microbiability and the heritability of dairy cattle CH4 production were largely independent. Thus, selective breeding for reduced CH4 production can be extended by methods perturbing the rumen microbiota towards reduced CH4 production. In Chapter 6 (the general discussion), the value of method comparisons for phenotype development by comparatively quantifying sources of error between cheaper alternative methods and intensive gold standard methods was discussed. The primary constraint to breeding for improved feed efficiency and CH4 production remains the recording of feed intake on a large scale under commercial conditions and recording of “true” CH4 production. It was proposed that the accuracy of bull breeding values for both feed efficiency and CH4 production can be increased through the use of sniffer phenotypes in robot milking herds, using individual level correlations but a genetic correlation between sniffer phenotypes and RC CH4 production are still needed. The records required for estimating genetic correlations with meaningful standard errors can only be achieved through substantial financial investments, development of cheaper alternative methods of phenotype recording or international collaborations. Further to the general discussion, a portion of host phenotypic variation in CH4 production was found to be associated with the rumen bacterial and archaeal composition. However, research is needed to determine if microbial associations are causative and methods to direct desired changes in the rumen microbial composition are still needed to unlock the potential of this under-exploited resource. The methods developed for quantifying the microbial contribution to host phenotypic variation will be of value to inform research into complex microbial-associated phenotypes, such as diseases and digestion in dairy cattle, other livestock species and humans. This thesis therefore contributes to the understanding of the genetic variation in feed efficiency, methane emissions and rumen metagenome of dairy cows.</p
Review: Genetic and genomic selection as a methane mitigation strategy in dairy cattle
Full text linkpublishedVersio
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
Dispelling the Myths Behind First-author Citation Counts
Full text linkWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
- …
