Crop design for specific adaptation in variable dryland production environments

Climatic variability in dryland production environments (E) generates variable yield and crop production risks. Optimal combinations of genotype (G) and management (M) depend strongly on E and thus vary among sites and seasons. Traditional crop improvement seeks broadly adapted genotypes to give best average performance under a standard management regime across the entire production region, with some subsequent manipulation of management regionally in response to average local environmental conditions. This process does not search the full spectrum of potential G × M × E combinations forming the adaptation landscape. Here we examine the potential value (relative to the conventional, broad adaptation approach) of exploiting specific adaptation arising from G × M × E. We present an in-silico analysis for sorghum production in Australia using the APSIM sorghum model. Crop design (G × M) is optimised for subsets of locations within the production region (specific adaptation) and is compared with the optimum G across all environments with locally modified M (broad adaptation). We find that geographic subregions that have frequencies of major environment types substantially different from that for the entire production region show greatest advantage for specific adaptation. Although the specific adaptation approach confers yield and production risk advantages at industry scale, even greater benefits should be achievable with better predictors of environment-type likelihood than that conferred by location alone.

Should I stay or should I go?: Rural and remote students in first year university stem courses

Research on the achievement of rural and remote students in science and mathematics is located within a context of falling levels of participation in physical science and mathematics courses in Australian schools, and underrepresentation of rural students in higher education. International studies such as the Programme of International Student Assessment (PISA), have reported lower levels of mathematical and scientific literacy in Australian students from rural and remote schools (Thomson et al, 2011). The SiMERR national survey of science, mathematics and ICT education in rural and regional Australia (Lyons et al, 2006) identified factors affecting student achievement in rural and remote schools. Many of the issues faced by rural and remote students in their schools are likely to have implications on their university enrolments in science, technology, engineering and mathematics (STEM) courses. For example, rural and remote students are less likely to attend university in general than their city counterparts and higher university attrition rates have been reported for remote students nationally. This paper examines the responses of a sample of rural/remote Australian first year STEM students at Australian universities to two questions. These related to their intentions to complete the course; and whether -and if so, why- they had ever considered withdrawing from their course. Results indicated that rural students who were still in their course by the end of first year were no more or less likely to consider withdrawing than were their peers from more populous centres. However, almost 20% of the rural cohort had considered withdrawing at some stage in their course, and their explanations provide insights into the reasoning of those who may not persist with their courses at university. These results, in the context of the greater attrition rate of remote students from university, point to the need to identify factors that positively impact on rural and remote students’ interest and achievement in science and mathematics. It also highlights a need for future research into the particular issues remote students may face in deciding whether or not to do science at the two key transition points of senior school and university/TAFE studies, and whether or not to persist in their tertiary studies. This paper is positioned at the intersection of two problems in Australian education. The first is a context of falling levels of participation in physical science and mathematics courses in Australian universities. The second is persistent inequitable access to, and retention in, tertiary education for students from rural and remote areas. Despite considerable research attention to both of these areas over recent years these problems have thus far proved to be intractable. This paper therefore aims to briefly review the relevant Australian literature pertaining to these issues; that is, declining STEM enrolments, and the underrepresentation and retention of rural/remote students in higher education. Given the related problems in these two overlapping domains, we then explore the views of first year rural students enrolled in courses, in relation to their intentions of withdrawing (or not) and the associated reasons for their views.

Genetic control of flowering in spotted gum, Corymbia citriodora subsp. variegata and C. maculata

Genetically controlled asynchrony in anthesis is an effective barrier to gene flow between planted and native forests. We investigated the degree of genetically controlled variation in the timing of key floral developmental stages in a major plantation species in subtropical Australia, Corymbia citriodora subsp. variegata K.D. Hill and L.A.S Johnson, and its relative C. maculata K.D. Hill and L.A.S. Johnson. Flowering observations were made in a common garden planting at Bonalbo in northern New South Wales in spring on 1855 trees from eight regions over three consecutive years, and monthly on a subset of 208 trees for 12 months. Peak anthesis time was stable over years and observations from translocated trees tended to be congruent with the observations in native stands, suggesting strong genetic control of anthesis time. A cluster of early flowering provenances was identified from the north-east of the Great Dividing Range. The recognition of a distinct flowering race from this region accorded well with earlier evidence of adaptive differentiation of populations from this region and geographically-structured genetic groupings in C. citriodora subsp. variegata. The early flowering northern race was more fecund, probably associated with its disease tolerance and greater vigour. Bud abundance fluctuated extensively at the regional level across 3 years suggesting bud abundance was more environmentally labile than timing of anthesis. Overall the level of flowering in the planted stand (age 12 years) was low (8–12% of assessed trees with open flowers), and was far lower than in nearby native stands. Low levels of flowering and asynchrony in peak anthesis between flowering races of C. citriodora subsp. variegata may partially mitigate a high likelihood of gene flow where the northern race is planted in the south of the species range neighbouring native stands

Weed dynamics as influenced by tillage system, sowing time and weed competition duration in dry-seeded rice

Cultural practices alter patterns of crop growth and can modify dynamics of weed-crop competition, and hence need to be investigated to evolve sustainable weed management in dry-seeded rice (DSR). Studies on weed dynamics in DSR sown at different times under two tillage systems were conducted at the Agronomic Research Farm, University of Agriculture, Faisalabad, Pakistan. A commonly grown fine rice cultivar 'Super Basmati' was sown on 15th June and 7th July of 2010 and 2011 under zero-till (ZT) and conventional tillage (CONT) and it was subjected to different durations of weed competition [10, 20, 30, 40, and 50 days after sowing (DAS) and season-long competition]. Weed-free plots were maintained under each tillage system and sowing time for comparison. Grassy weeds were higher under ZT while CONT had higher relative proportion of broad-leaved weeds in terms of density and biomass. Density of sedges was higher by 175% in the crop sown on the 7th July than on the 15th June. Delaying sowing time of DSR from mid June to the first week of July reduced weed density by 69 and 43% but their biomass remained unaffected. Tillage systems had no effect on total weed biomass. Plots subjected to season-long weed competition had mostly grasses while broad-leaved weeds were not observed at harvest. In the second year of study, dominance of grassy weeds was increased under both tillage systems and sowing times. Significantly less biomass (48%) of grassy weeds was observed under CONT than ZT in 2010; however, during 2011, this effect was non-significant. Trianthema portulacastrum and Dactyloctenium aegyptium were the dominant broad-leaved and grassy weeds, respectively. Cyperus rotundus was the dominant sedge weed, especially in the crop sown on the 7th July. Relative yield loss (RYL) ranged from 3 to 13% and 7 to16% when weeds were allowed to compete only for 20 DAS. Under season-long weed competition, RYL ranged from 68 to 77% in 2010 and 74 to80% in 2011. The sowing time of 15th June was effective in minimizing weed proliferation and rectifying yield penalty associated with the 7th July sowing. The results suggest that DSR in Pakistan should preferably be sown on 15th June under CONT systems and weeds must be controlled before 20 DAS to avoid yield losses. Successful adoption of DSR at growers' fields in Pakistan will depend on whether growers can control weeds and prevent shifts in weed population from intractable weeds to more difficult-to-control weeds as a consequence of DSR adoption.

Weeds of Direct-Seeded Rice in Asia: Problems and Opportunities

Rice production symbolizes the single largest land use for food production on the Earth. The significance of this cereal as a source of energy and income seems overwhelming for millions of people in Asia, representing 90% of global rice production and consumption. Estimates indicate that the burgeoning population will need 25% more rice by 2025 than today's consumption. As the demand for rice is increasing, its production in Asia is threatened by a dwindling natural resource base, socioeconomic limitations, and uncertainty of climatic optima. Transplanting in puddled soil with continuous flooding is a common method of rice crop establishment in Asia. There is a dire need to look for rice production technologies that not only cope with existing limitations of transplanted rice but also are viable, economical, and secure for future food demand.Direct seeding of rice has evolved as a potential alternative to the current detrimental practice of puddling and nursery transplanting. The associated benefits include higher water productivity, less labor and energy inputs, less methane emissions, elimination of time and edaphic conflicts in the rice-wheat cropping system, and early crop maturity. Realization of the yield potential and sustainability of this resource-conserving rice production technique lies primarily in sustainable weed management, since weeds have been recognized as the single largest biological constraint in direct-seeded rice (DSR). Weed competition can reduce DSR yield by 30-80% and even complete crop failure can occur under specific conditions. Understanding the dynamics and outcomes of weed-crop competition in DSR requires sound knowledge of weed ecology, besides production factors that influence both rice and weeds, as well as their association. Successful adoption of direct seeding at the farmers' level in Asia will largely depend on whether farmers can control weeds and prevent shifts in weed populations from intractable weeds to more difficult-to-control weeds as a consequence of direct seeding. Sustainable weed management in DSR comprises all the factors that give DSR a competitive edge over weeds regarding acquisition and use of growth resources. This warrants the need to integrate various cultural practices with weed control measures in order to broaden the spectrum of activity against weed flora. A weed control program focusing entirely on herbicides is no longer ecologically sound, economically feasible, and effective against diverse weed flora and may result in the evolution of herbicide-resistant weed biotypes. Rotation of herbicides with contrasting modes of action in conjunction with cultural measures such as the use of weed-competitive rice cultivars, sowing time, stale seedbed technique, seeding rate, crop row spacing, fertilizer and water inputs and their application method/timing, and manual and mechanical hoeing can prove more effective and need to be optimized keeping in view the type and intensity of weed infestation. This chapter tries to unravel the dynamics of weed-crop competition in DSR. Technological issues, limitations associated with DSR, and opportunities to combat the weed menace are also discussed as a pragmatic approach for sustainable DSR production. A realistic approach to secure yield targets against weed competition will combine the abovementioned strategies and tactics in a coordinated manner. This chapter further suggests the need of multifaceted and interdisciplinary research into ecologically based weed management, as DSR seems inevitable in the near future.

Enhanced processing of SPOT multispectral satellite imagery for environmental monitoring and modelling

The Taita Hills in southeastern Kenya form the northernmost part of Africa’s Eastern Arc Mountains, which have been identified by Conservation International as one of the top ten biodiversity hotspots on Earth. As with many areas of the developing world, over recent decades the Taita Hills have experienced significant population growth leading to associated major changes in land use and land cover (LULC), as well as escalating land degradation, particularly soil erosion. Multi-temporal medium resolution multispectral optical satellite data, such as imagery from the SPOT HRV, HRVIR, and HRG sensors, provides a valuable source of information for environmental monitoring and modelling at a landscape level at local and regional scales. However, utilization of multi-temporal SPOT data in quantitative remote sensing studies requires the removal of atmospheric effects and the derivation of surface reflectance factor. Furthermore, for areas of rugged terrain, such as the Taita Hills, topographic correction is necessary to derive comparable reflectance throughout a SPOT scene. Reliable monitoring of LULC change over time and modelling of land degradation and human population distribution and abundance are of crucial importance to sustainable development, natural resource management, biodiversity conservation, and understanding and mitigating climate change and its impacts. The main purpose of this thesis was to develop and validate enhanced processing of SPOT satellite imagery for use in environmental monitoring and modelling at a landscape level, in regions of the developing world with limited ancillary data availability. The Taita Hills formed the application study site, whilst the Helsinki metropolitan region was used as a control site for validation and assessment of the applied atmospheric correction techniques, where multiangular reflectance field measurements were taken and where horizontal visibility meteorological data concurrent with image acquisition were available. The proposed historical empirical line method (HELM) for absolute atmospheric correction was found to be the only applied technique that could derive surface reflectance factor within an RMSE of < 0.02 ps in the SPOT visible and near-infrared bands; an accuracy level identified as a benchmark for successful atmospheric correction. A multi-scale segmentation/object relationship modelling (MSS/ORM) approach was applied to map LULC in the Taita Hills from the multi-temporal SPOT imagery. This object-based procedure was shown to derive significant improvements over a uni-scale maximum-likelihood technique. The derived LULC data was used in combination with low cost GIS geospatial layers describing elevation, rainfall and soil type, to model degradation in the Taita Hills in the form of potential soil loss, utilizing the simple universal soil loss equation (USLE). Furthermore, human population distribution and abundance were modelled with satisfactory results using only SPOT and GIS derived data and non-Gaussian predictive modelling techniques. The SPOT derived LULC data was found to be unnecessary as a predictor because the first and second order image texture measurements had greater power to explain variation in dwelling unit occurrence and abundance. The ability of the procedures to be implemented locally in the developing world using low-cost or freely available data and software was considered. The techniques discussed in this thesis are considered equally applicable to other medium- and high-resolution optical satellite imagery, as well the utilized SPOT data.

On probability-based inference under data missing by design

Whether a statistician wants to complement a probability model for observed data with a prior distribution and carry out fully probabilistic inference, or base the inference only on the likelihood function, may be a fundamental question in theory, but in practice it may well be of less importance if the likelihood contains much more information than the prior. Maximum likelihood inference can be justified as a Gaussian approximation at the posterior mode, using flat priors. However, in situations where parametric assumptions in standard statistical models would be too rigid, more flexible model formulation, combined with fully probabilistic inference, can be achieved using hierarchical Bayesian parametrization. This work includes five articles, all of which apply probability modeling under various problems involving incomplete observation. Three of the papers apply maximum likelihood estimation and two of them hierarchical Bayesian modeling. Because maximum likelihood may be presented as a special case of Bayesian inference, but not the other way round, in the introductory part of this work we present a framework for probability-based inference using only Bayesian concepts. We also re-derive some results presented in the original articles using the toolbox equipped herein, to show that they are also justifiable under this more general framework. Here the assumption of exchangeability and de Finetti's representation theorem are applied repeatedly for justifying the use of standard parametric probability models with conditionally independent likelihood contributions. It is argued that this same reasoning can be applied also under sampling from a finite population. The main emphasis here is in probability-based inference under incomplete observation due to study design. This is illustrated using a generic two-phase cohort sampling design as an example. The alternative approaches presented for analysis of such a design are full likelihood, which utilizes all observed information, and conditional likelihood, which is restricted to a completely observed set, conditioning on the rule that generated that set. Conditional likelihood inference is also applied for a joint analysis of prevalence and incidence data, a situation subject to both left censoring and left truncation. Other topics covered are model uncertainty and causal inference using posterior predictive distributions. We formulate a non-parametric monotonic regression model for one or more covariates and a Bayesian estimation procedure, and apply the model in the context of optimal sequential treatment regimes, demonstrating that inference based on posterior predictive distributions is feasible also in this case.

Joint Regression and Association Models for Repeated Categorical Responses

The focus of this study is on statistical analysis of categorical responses, where the response values are dependent of each other. The most typical example of this kind of dependence is when repeated responses have been obtained from the same study unit. For example, in Paper I, the response of interest is the pneumococcal nasopharengyal carriage (yes/no) on 329 children. For each child, the carriage is measured nine times during the first 18 months of life, and thus repeated respones on each child cannot be assumed independent of each other. In the case of the above example, the interest typically lies in the carriage prevalence, and whether different risk factors affect the prevalence. Regression analysis is the established method for studying the effects of risk factors. In order to make correct inferences from the regression model, the associations between repeated responses need to be taken into account. The analysis of repeated categorical responses typically focus on regression modelling. However, further insights can also be gained by investigating the structure of the association. The central theme in this study is on the development of joint regression and association models. The analysis of repeated, or otherwise clustered, categorical responses is computationally difficult. Likelihood-based inference is often feasible only when the number of repeated responses for each study unit is small. In Paper IV, an algorithm is presented, which substantially facilitates maximum likelihood fitting, especially when the number of repeated responses increase. In addition, a notable result arising from this work is the freely available software for likelihood-based estimation of clustered categorical responses.

Flying-Fox Roost Disturbance and Hendra Virus Spillover Risk

Bats of the genus Pteropus (flying-foxes) are the natural host of Hendra virus (HeV) which periodically causes fatal disease in horses and humans in Australia. The increased urban presence of flying-foxes often provokes negative community sentiments because of reduced social amenity and concerns of HeV exposure risk, and has resulted in calls for the dispersal of urban flying-fox roosts. However, it has been hypothesised that disturbance of urban roosts may result in a stress-mediated increase in HeV infection in flying-foxes, and an increased spillover risk. We sought to examine the impact of roost modification and dispersal on HeV infection dynamics and cortisol concentration dynamics in flying-foxes. The data were analysed in generalised linear mixed models using restricted maximum likelihood (REML). The difference in mean HeV prevalence in samples collected before (4.9%), during (4.7%) and after (3.4%) roost disturbance was small and non-significant (P = 0.440). Similarly, the difference in mean urine specific gravity-corrected urinary cortisol concentrations was small and non-significant (before = 22.71 ng/mL, during = 27.17, after = 18.39) (P= 0.550). We did find an underlying association between cortisol concentration and season, and cortisol concentration and region, suggesting that other (plausibly biological or environmental) variables play a role in cortisol concentration dynamics. The effect of roost disturbance on cortisol concentration approached statistical significance for region, suggesting that the relationship is not fixed, and plausibly reflecting the nature and timing of disturbance. We also found a small positive statistical association between HeV excretion status and urinary cortisol concentration. Finally, we found that the level of flying-fox distress associated with roost disturbance reflected the nature and timing of the activity, highlighting the need for a ‘best practice’ approach to dispersal or roost modification activities. The findings usefully inform public discussion and policy development in relation to Hendra virus and flying-fox management.

Hazard Function Models to Estimate Mortality Rates Affecting Fish Populations with Application to the Sea Mullet (Mugil cephalus) Fishery on the Queensland Coast (Australia)

Fisheries management agencies around the world collect age data for the purpose of assessing the status of natural resources in their jurisdiction. Estimates of mortality rates represent a key information to assess the sustainability of fish stocks exploitation. Contrary to medical research or manufacturing where survival analysis is routinely applied to estimate failure rates, survival analysis has seldom been applied in fisheries stock assessment despite similar purposes between these fields of applied statistics. In this paper, we developed hazard functions to model the dynamic of an exploited fish population. These functions were used to estimate all parameters necessary for stock assessment (including natural and fishing mortality rates as well as gear selectivity) by maximum likelihood using age data from a sample of catch. This novel application of survival analysis to fisheries stock assessment was tested by Monte Carlo simulations to assert that it provided unbiased estimations of relevant quantities. The method was applied to the data from the Queensland (Australia) sea mullet (Mugil cephalus) commercial fishery collected between 2007 and 2014. It provided, for the first time, an estimate of natural mortality affecting this stock: 0.22±0.08 year −1 .

Computing the Stochastic Complexity of Simple Probabilistic Graphical Models

Minimum Description Length (MDL) is an information-theoretic principle that can be used for model selection and other statistical inference tasks. There are various ways to use the principle in practice. One theoretically valid way is to use the normalized maximum likelihood (NML) criterion. Due to computational difficulties, this approach has not been used very often. This thesis presents efficient floating-point algorithms that make it possible to compute the NML for multinomial, Naive Bayes and Bayesian forest models. None of the presented algorithms rely on asymptotic analysis and with the first two model classes we also discuss how to compute exact rational number solutions.

Computationally Efficient Methods for MDL-Optimal Density Estimation and Data Clustering

The Minimum Description Length (MDL) principle is a general, well-founded theoretical formalization of statistical modeling. The most important notion of MDL is the stochastic complexity, which can be interpreted as the shortest description length of a given sample of data relative to a model class. The exact definition of the stochastic complexity has gone through several evolutionary steps. The latest instantation is based on the so-called Normalized Maximum Likelihood (NML) distribution which has been shown to possess several important theoretical properties. However, the applications of this modern version of the MDL have been quite rare because of computational complexity problems, i.e., for discrete data, the definition of NML involves an exponential sum, and in the case of continuous data, a multi-dimensional integral usually infeasible to evaluate or even approximate accurately. In this doctoral dissertation, we present mathematical techniques for computing NML efficiently for some model families involving discrete data. We also show how these techniques can be used to apply MDL in two practical applications: histogram density estimation and clustering of multi-dimensional data.

Quantification of the inevitable: the influence of soil macrofauna on soil water movement in rehabilitated open-cut mined lands

Recolonisation of soil by macrofauna (especially ants, termites and earthworms) in rehabilitated open-cut mine sites is inevitable and, in terms of habitat restoration and function, typically of great value. In these highly disturbed landscapes, soil invertebrates play a major role in soil development (macropore configuration, nutrient cycling, bioturbation, etc.) and can influence hydrological processes such as infiltration, seepage, runoff generation and soil erosion. Understanding and quantifying these ecosystem processes is important in rehabilitation design, establishment and subsequent management to ensure progress to the desired end goal, especially in waste cover systems designed to prevent water reaching and transporting underlying hazardous waste materials. However, the soil macrofauna is typically overlooked during hydrological modelling, possibly due to uncertainties on the extent of their influence, which can lead to failure of waste cover systems or rehabilitation activities. We propose that scientific experiments under controlled conditions and field trials on post-mining lands are required to quantify (i) macrofauna–soil structure interactions, (ii) functional dynamics of macrofauna taxa,and (iii) their effects on macrofauna and soil development over time. Such knowledge would provide crucial information for soil water models, which would increase confidence in mine waste cover design recommendations and eventually lead to higher likelihood of rehabilitation success of open-cut mining land.

Co2 emissions and income trajectory in Australia: The role of technological change

This study investigates the relationship between per capita carbon dioxide (CO2) emissions and per capita GDP in Australia, while controlling for technological state as measured by multifactor productivity and export of black coal. Although technological progress seems to play a critical role in achieving long term goals of CO2 reduction and economic growth, empirical studies have often considered time trend to proxy technological change. However, as discoveries and diffusion of new technologies may not progress smoothly with time, the assumption of a deterministic technological progress may be incorrect in the long run. The use of multifactor productivity as a measure of technological state, therefore, overcomes the limitations and provides practical policy directions. This study uses recently developed bound-testing approach, which is complemented by Johansen- Juselius maximum likelihood approach and a reasonably large sample size to investigate the cointegration relationship. Both of the techniques suggest that cointegration relationship exists among the variables. The long-run and short-run coefficients of CO2 emissions function is estimated using ARDL approach. The empirical findings in the study show evidence of the existence of Environmental Kuznets Curve type relationship for per capita CO2 emissions in the Australian context. The technology as measured by the multifactor productivity, however, is not found as an influencing variable in emissionsincome trajectory.

Head motion and Inattention/Hyperactivity share common genetic influences: Implications for fMRI studies of ADHD

Head motion (HM) is a well known confound in analyses of functional MRI (fMRI) data. Neuroimaging researchers therefore typically treat HM as a nuisance covariate in their analyses. Even so, it is possible that HM shares a common genetic influence with the trait of interest. Here we investigate the extent to which this relationship is due to shared genetic factors, using HM extracted from resting-state fMRI and maternal and self report measures of Inattention and Hyperactivity-Impulsivity from the Strengths and Weaknesses of ADHD Symptoms and Normal Behaviour (SWAN) scales. Our sample consisted of healthy young adult twins (N = 627 (63% females) including 95 MZ and 144 DZ twin pairs, mean age 22, who had mother-reported SWAN; N = 725 (58% females) including 101 MZ and 156 DZ pairs, mean age 25, with self reported SWAN). This design enabled us to distinguish genetic from environmental factors in the association between head movement and ADHD scales. HM was moderately correlated with maternal reports of Inattention (r = 0.17, p-value = 7.4E-5) and Hyperactivity-Impulsivity (r = 0.16, p-value = 2.9E-4), and these associations were mainly due to pleiotropic genetic factors with genetic correlations [95% CIs] of rg = 0.24 [0.02, 0.43] and rg = 0.23 [0.07, 0.39]. Correlations between self-reports and HM were not significant, due largely to increased measurement error. These results indicate that treating HM as a nuisance covariate in neuroimaging studies of ADHD will likely reduce power to detect between-group effects, as the implicit assumption of independence between HM and Inattention or Hyperactivity-Impulsivity is not warranted. The implications of this finding are problematic for fMRI studies of ADHD, as failing to apply HM correction is known to increase the likelihood of false positives. We discuss two ways to circumvent this problem: censoring the motion contaminated frames of the RS-fMRI scan or explicitly modeling the relationship between HM and Inattention or Hyperactivity-Impulsivity