Inferential, non-parametric statistics to assess quality of probabilistic forecast systems

Many statistical forecast systems are available to interested users. In order to be useful for decision-making, these systems must be based on evidence of underlying mechanisms. Once causal connections between the mechanism and their statistical manifestation have been firmly established, the forecasts must also provide some quantitative evidence of `quality’. However, the quality of statistical climate forecast systems (forecast quality) is an ill-defined and frequently misunderstood property. Often, providers and users of such forecast systems are unclear about what ‘quality’ entails and how to measure it, leading to confusion and misinformation. Here we present a generic framework to quantify aspects of forecast quality using an inferential approach to calculate nominal significance levels (p-values) that can be obtained either by directly applying non-parametric statistical tests such as Kruskal-Wallis (KW) or Kolmogorov-Smirnov (KS) or by using Monte-Carlo methods (in the case of forecast skill scores). Once converted to p-values, these forecast quality measures provide a means to objectively evaluate and compare temporal and spatial patterns of forecast quality across datasets and forecast systems. Our analysis demonstrates the importance of providing p-values rather than adopting some arbitrarily chosen significance levels such as p < 0.05 or p < 0.01, which is still common practice. This is illustrated by applying non-parametric tests (such as KW and KS) and skill scoring methods (LEPS and RPSS) to the 5-phase Southern Oscillation Index classification system using historical rainfall data from Australia, The Republic of South Africa and India. The selection of quality measures is solely based on their common use and does not constitute endorsement. We found that non-parametric statistical tests can be adequate proxies for skill measures such as LEPS or RPSS. The framework can be implemented anywhere, regardless of dataset, forecast system or quality measure. Eventually such inferential evidence should be complimented by descriptive statistical methods in order to fully assist in operational risk management.

Near infrared reflectance as a rapid and inexpensive surrogate measure for fatty acid composition and oil content in peanuts (Arachis hypogaea L.)

The fatty acid composition of ground nuts (Arachis hypogaea L.) commonly known as peanuts, is an important consideration when a new variety is being released. The composition impacts on nutrition and, importantly, self-life of peanut products. To select for suitable breeding material, it was necessary to develop a rapid, non-derstructive and cost-efficient method. Near infrared spectroscopy was chosen as that methodology. Calibrations were developed for two major fatty-acid components, oleic and linoleic acids and two minor components, palmitic and stearic acids, as well as total oil content. Partial least squares models indicated a high level of precision with a squared multiple correlation coefficient of greater than 0.90 for each constitutent. Standard errors for prediction for oleic, linoleic, palmitic, stearic acids and total oil content were 6.4%, 4.5%, 0.8%, 0.9% and 1.3% respectively. The results demonstrated that reasonable calibrations could be developed to predict oil composition and content of peanuts for a breeding programme.

Impact of resistance on the efficacy of binary combinations of spinosad, chlorpyrifos-methyl and s-methoprene against five stored-grain beetles

Laboratory experiments were conducted to determine the efficacy of spinosad (a biopesticide), chlorpyrifos-methyl (an organophosphorus compound (OP)) and s-methoprene (a juvenile hormone analogue) applied alone and in binary combinations against five stored-grain beetles in wheat. There were three strains of Rhyzopertha dominica, and one strain each of Sitophilus oryzae, Tribolium castaneum, Oryzaephilus surinamensis and Cryptolestes ferrugineus. These strains were chosen to represent a range of possible resistant genotypes, exhibiting resistance to organophosphates, pyrethroids or methoprene. Treatments were applied at rates that are registered or likely to be registered in Australia. Adults were exposed to freshly treated wheat for 2 weeks, and the effects of treatments on mortality and reproduction were determined. No single protectant or protectant combination controlled all insect strains, based on the criterion of >99% reduction in the number of live F1 adults relative to the control. The most effective combinations were spinosad at 1 mg kg-1+chlorpyrifos-methyl at 10 mg kg-1 which controlled all strains except for OP-resistant O. surinamensis, and chlorpyrifos-methyl at 10 mg kg-1+s-methoprene at 0.6 mg kg-1 which controlled all strains except for methoprene-resistant R. dominica. The results of this study demonstrate the difficulty in Australia, and potentially other countries which use protectants, of finding protectant treatments to control a broad range of pest species in the face of resistance development.

Influence of food supplementation on the fitness of two biological control agents: a predatory nabid bug and a bollworm pupal parasitoid

Many arthropod predators and parasitoids exhibit either stage-specific or lifetime omnivory, in that they include extra-floral nectar, floral nectar, honeydew or pollen in their immature and/or adult diet. Access to these plant-derived foods can enhance pest suppression by increasing both the individual fitness and local density of natural enemies. Commercial products such as Amino-Feed®, Envirofeast®, and Pred-Feed® can be applied to crops to act as artificial-plant-derived foods. In laboratory and glasshouse experiments we examined the influence of carbohydrate and protein rich Amino-Feed UV® or Amino-Feed, respectively, on the fitness of a predatory nabid bug Nabis kinbergii Reuter (Hemiptera: Nabidae) and bollworm pupal parasitoid Ichneumon promissorius (Erichson) (Hymenoptera: Ichneumonidae). Under the chosen conditions, the provision of either wet or dry residues of Amino-Feed UV had no discernable effect on immediate or longer-term survival and immature development times of N. kinbergii. In contrast, the provision of honey, Amino-Feed plus extrafloral nectar, and extrafloral nectar alone had a marked effect on the longevity of I. promissorius, indicating that they were limited by at least carbohydrates as an energy source, but probably not protein. Compared with a water only diet, the provision of Amino-Feed plus extrafloral nectar increased the longevity of males and females of I. promissorius by 3.0- and 2.4-fold, respectively. Not only did female parasitoids live longer when provided food, but the total number of eggs laid and timing of deposition was affected by diet under the chosen conditions. Notably, females in the water and honey treatments deposited greater numbers of eggs earlier in the trial, but this trend was unable to be sustained over their lifetime. Egg numbers in these treatments subsequently fell below the levels achieved by females in the Amino-Feed plus extrafloral nectar and cotton extrafloral nectar only treatments. Furthermore, there were times when the inclusion of the Amino-Feed was beneficial compared with cotton extrafloral nectar only. Artificial food supplements and plant-derived foods are worthy of further investigation because they have potential to improve the ecosystem service of biological pest control in targeted agroecosystems by providing natural enemies with an alternative source of nutrition, particularly during periods of prey/host scarcity.

Rapid and sensitive detection of Avibacterium paragallinarum in the presence of other bacteria using a 5' Taq nuclease assay: A new tool for diagnosing infectious coryza

A 5' Taq nuclease assay specific for Avibacterium paragallinarum was designed and optimized for use in diagnosing infectious coryza. The region chosen for assay design was one of known specificity for Av. paragallinarum. The assay detected Av. paragallinarum reference strains representing the three Page and the eight Kume serovars, and field isolates from diverse geographical locations. No cross-reactions were observed with other Avibacterium species, with other bacteria taxonomically related to Av. paragallinarum nor with bacteria and viruses likely to be present in swabs collected from suspected infectious coryza cases. The detection limit for the assay was 6 to 60 colony-forming units per reaction. Twenty-two out of 53 swabs collected from sick birds reacted in the 5' Taq nuclease assay, whereas Av. paragallinarum was not isolated from any of the swabs. All of the 22 swabs yielded other bacteria in culture. The presence of Av. paragallinarum in the swabs was also demonstrated by sequencing, thereby confirming the ability of the assay to detect Av. paragallinarum in the presence of other bacteria. The ability to quantify bacterial load in the swabs using the 5' Taq nuclease assay was demonstrated.

Gene expression in the skin of Bos taurus and Bos indicus cattle infested with the cattle tick, Rhipicephalus (Boophilus) microplus

The cattle tick Rhipicephalus microplus (formerly Boophilus microplus) is responsible for severe production losses to the cattle industry worldwide. It has long been known that different breeds of cattle can resist tick infestation to varying degrees; however, the mechanisms by which resistant cattle prevent heavy infestation are largely unknown. The aim of this study was to determine whether gene expression varied significantly between skin sampling sites (neck, chest and tail region), and whether changes in gene expression could be detected in samples taken at tick attachment sites (tick attached to skin sample) compared with samples taken from non-attachment sites (no tick attachment). We present here the results of an experiment examining the expression of a panel of forty-four genes in skin sections taken from Bos indicus (Brahman) cattle of known high resistance, and Bos taurus (Holstein-Friesian) cattle of known low resistance to the cattle tick. The forty-four genes chosen for this study included genes known to be involved in several immune processes, some structural genes, and some genes previously suggested to be of importance in tick resistance by other researchers. The expression of fifteen gene transcripts increased significantly in Holstein-Friesian skin samples at tick attachment sites. The higher expression of many genes involved in innate inflammatory processes in the Holstein-Friesian animals at tick attachment sites suggests this breed is exhibiting a non-directed pathological response to infestation. Of the forty-four genes analysed, no transcripts were detected in higher abundance at tick attachment sites in the Brahman cattle compared with similar samples from the Holstein-Friesian group, nor difference between attachment site and non-attachment site samples within the Brahman group. The results presented here suggest that the means by which these two cattle breeds respond to tick infestation differ and warrant further investigation.

NIR model development and robustness in prediction of melon fruit total soluble solids

Near infrared spectroscopy (NIRS) can be used for the on-line, non-invasive assessment of fruit for eating quality attributes such as total soluble solids (TSS). The robustness of multivariate calibration models, based on NIRS in a partial transmittance optical geometry, for the assessment of TSS of intact rockmelons (Cucumis melo) was assessed. The mesocarp TSS was highest around the fruit equator and increased towards the seed cavity. Inner mesocarp TSS levels decreased towards both the proximal and distal ends of the fruit, but more so towards the proximal end. The equatorial region of the fruit was chosen as representative of the fruit for near infrared assessment of TSS. The spectral window for model development was optimised at 695-1045 nm, and the data pre-treatment procedure was optimised to second-derivative absorbance without scatter correction. The 'global' modified partial least squares (MPLS) regression modelling procedure of WINISI (ver. 1.04) was found to be superior with respect to root mean squared error of prediction (RMSEP) and bias for model predictions of TSS across seasons, compared with the 'local' MPLS regression procedure. Updating of the model with samples selected randomly from the independent validation population demonstrated improvement in both RMSEP and bias with addition of approximately 15 samples.

Response of perennial ryegrass (Lolium perenne) to renovation in Australian dairy pastures

This study reports on the effect of oversowing perennial ryegrass (Lolium perenne L.) into a degraded perennial ryegrass and white clover (Trifolium repens L.) pasture to extend its productive life using various intensities of seedbed preparation. Sites in New South Wales (NSW), Western Australia (WA), South Australia (SA) and Tasmania (Tas.) were chosen by a local group of farmers as being degraded and in need of renovation. Control (nil renovation) and medium (mulch and graze, spray with glyphosphate and sow) renovation treatments were common to all sites whereas minimum (mulch and graze, and sow) and full seedbed (graze and spray with glyphosphate and then full seedbed preparation) renovation were imposed only at some sites. Plots varied in area from 0.14 to 0.50 ha, and were renovated then sown in March or April 2000 and subsequently grazed by dairy cows. Pasture utilisation was estimated from pre- and post-grazing pasture mass assessed by a rising plate pasture meter. Utilised herbage mass of the renovated treatments was significantly higher than control plots in period 1 (planting to August) and 2 (first spring) at the NSW site only. There was no difference among treatments in period 3 (first summer) at any site, and only at the WA and NSW sites in period 4 (March to July 2001) was there a response to renovation. As a result, renovation at the NSW site only significantly increased ryegrass utilisation over the whole experimental period. Ryegrass plant density was higher at the NSW, WA (excluding minimum renovation) and Tas. (excluding full renovation) sites 6 months after renovation but this was only sustained for 12 months for the minimum and medium treatments at the NSW and Tas. sites, respectively, presumably due to reduced competition from naturalised C4 summer grasses [kikuyu (Pennisetum clandestinum) and paspalum (Paspalum dilatatum)] in NSW At the NSW, WA and SA sites, the original ryegrass plant density was low (<35 plants/m2) compared with the Tas. site where density was around 185/m2. The response to renovating a degraded perennial ryegrass pasture varied between sites in Australia. Positive responses were generally small and were most consistent where renovation removed competing C4 summer grasses.

Accelerated drying of plantation grown Eucalyptus cloeziana and Eucalyptus pellita sawn timber

The objective of this study was to gain an understanding for drying sawn timber produced from fast-grown, well-managed Queensland hardwood plantations using accelerated drying methods. Due to limited resources, this was a preliminary study and further work will be required to optimize schedules for industrial implementation. Three conventional kiln trials, including two for 38-mm-thick, 19-year-old plantation Gympie messmate (Eucalyptus cloeziana F. Muell.) and one for 25mm thick, 15-year-old plantation red mahogany (Eucalyptus pellita F. Muell.), and two vacuum kiln drying trials, one each for 38- and 25mm thick Gympie messmate, were conducted. Measurements of final cross-sectional moisture content, moisture content gradient, residual drying stress, and internal and surface checking were used to quantify dried quality. Drying schedules were chosen based on either existing published schedules or, in the case of the vacuum drying trials, existing schedules for species with similar wood density and dying degrade properties, or manipulated schedules based on the results of trials conducted during this study. The findings indicate that both species can be dried using conventional drying techniques with acceptable grade quality in approximately 75 percent of the drying time that industry is currently achieving when drying native forest timber of the same species. The vacuum drying time was 60 percent less than conventional drying for 38-mm-thick, 19-year-old Gympie messmate, although drying quality needs improving. The findings have shown that through careful schedule manipulation and adjustment, the grade quality can be optimized to suit the desired expectation. Additional research is required to further optimize the schedules to ensure acceptable grade qualities can be reliably achieved across all drying criteria and exploit opportunities to reduce drying times further.

Climate Change - Risks and Opportunities for the Custard Apple Industry

Climate affects the custard apple industry in a range of ways through impacts on growth, disease risk, fruit set and industry location. Climates in Australia are influenced by surrounding oceans, and are very variable from year to year. However, amidst this variability there are significant trends, with Australian annual mean temperatures increasing since 1910, and particularly since 1950, with night-time temperatures increasing faster (0.11oC/decade) than daytime temperatures (0.06oC/decade). These temperature increases and other climate changes are expected to continue as a result of greenhouse gas emissions, with ongoing impacts on the custard apple industry. Five sites were chosen to assess possible future climate changes : Mareeba, Yeppoon, Bundaberg, Nambour and Lismore, these sites representing the extent of the majority of custard apple production in eastern Australia. A fifth site (Coffs Harbour) was selected as it is south of the current production regions. A mean warming of 0.8 to 1.2oC is anticipated over most of these sites by the year 2030, relative to 1990. This paper assesses the potential effects of climate change on custard apple production, and suggests strategies for adaptation.

Commercial products from bio-active extractives in cypress milling residues.

Extractive components obtained from milling residues of white cypress were studied for chemical identity and bioactivity with a view to developing a commercial use for these components, thus increasing the value of the residues and improving the economics of cypress sawn wood production. Extracts obtained by solvent or steam extraction techniques from cypress sawdust were each fractionated by a range of techniques into groups of similar compounds. Crude extracts and fractions were screened against a range of agricultural pests and diseases, including two fungi, subterranean termites, fruit spotting bugs, two-spotted mites, thrips, heliothis, banana scab moths, silverleaf whiteflies, cattle tick adults and larvae, and ruminant gastrointestinal nematodes. Additional screening was undertaken where encouraging results were achieved, for two-spotted mites, thrips, silverleaf whiteflies, cattle tick adults and ruminant gastrointestinal nematodes. After considering degrees of efficacy against, and economic importance of, the agricultural pests, and likely production costs of extracts and fractions, the crude extract (oil) produced by steam distillation was chosen for further study against silverleaf whitefly. A useful degree of control was achievable when this oil was applied to tomato or eggplant at 0.1%, with much less harmful effects on a beneficial insect. Activity of the oil against silverleaf whitefly was undiminished 3.5 years after it was generated. There was little benefit from supplementing the extract with co-formulated paraffinic oil. From the steam distilled oil, fifty-five compounds were characterised, thirty-five compounds representing 92.478 % of the oil, with guaiol (20.8%) and citronellic acid (15.9%) most abundant. These two compounds, and a group of oxygenated compounds containing bulnesol and a range of eudesmols, were found to account for most of the activity against silverleaf whitefly. This application was recommended for first progression to commercialisation.

Multi-trait assessment of early-in-life female, male and genomic measures for use in genetic selection to improve female reproductive performance of Brahman cattle

Early-in-life female and male measures with potential to be practical genetic indicators were chosen from earlier analyses and examined together with genomic measures for multi-trait use to improve female reproduction of Brahman cattle. Combinations of measures were evaluated on the genetic gains expected from selection of sires and dams for each of age at puberty (AGECL, i.e. first observation of a corpus luteum), lactation anoestrous interval in 3-year-old cows (LAI), and lifetime annual weaning rate (LAWR, i.e. the weaning rate of cows based on the number of annual matings they experienced over six possible matings). Selection was on an index of comparable records for each combination. Selection intensities were less than theoretically possible but assumed a concerted selection effort was able to be made across the Brahman breed. The results suggested that substantial genetic gains could be possible but need to be confirmed in other data. The estimated increase in LAWR in 10 years, for combinations without or with genomic measures, ranged from 8 to 12 calves weaned per 100 cows from selection of sires, and from 12 to 15 calves weaned per 100 cows from selection of sires and dams. Corresponding reductions in LAI were 60-103 days or 94-136 days, and those for AGECL were 95-125 or 141-176 days, respectively. Coat score (a measure of the sleekness or wooliness of the coat) and hip height in females, and preputial eversion and liveweight in males, were measures that may warrant wider recording for Brahman female reproduction genetic evaluation. Pregnancy-test outcomes from Matings 1 and 2 also should be recorded. Percentage normal sperm may be important to record for reducing LAI and scrotal size and serum insulin-like growth factor-I concentration in heifers at 18 months for reducing AGECL. Use of a genomic estimated breeding value (EBV) in combination with other measures added to genetic gains, especially at genomic EBV accuracies of 40%. Accuracies of genomic EBVs needed to approach 60% for the genomic EBV to be the most important contributor to gains in the combinations of measures studied.

Floristic composition and pasture condition of Aristida/Bothriochloa pastures in central Queensland. II. Soil and pasture condition interactions

Sustainable management of native pastures requires an understanding of what the bounds of pasture composition, cover and soil surface condition are for healthy pastoral landscapes to persist. A survey of 107 Aristida/Bothriochloa pasture sites in inland central Queensland was conducted. The sites were chosen for their current diversity of tree cover, apparent pasture condition and soil type to assist in setting more objective bounds on condition ‘states’ in such pastures. Assessors’ estimates of pasture condition were strongly correlated with herbage mass (r = 0.57) and projected ground cover (r = 0. 58), and moderately correlated with pasture crown cover (r = 0.35) and tree basal area (r = 0.32). Pasture condition was not correlated with pasture plant density or the frequency of simple guilds of pasture species. The soil type of Aristida/Bothriochloa pasture communities was generally hard-setting, low in cryptogam cover but moderately covered with litter and projected ground cover (30–50%). There was no correlation between projected ground cover of pasture and estimated ground-level cover of plant crowns. Tree basal area was correlated with broad categories of soil type, probably because greater tree clearing has occurred on the more fertile, heavy-textured clay soils. Of the main perennial grasses, some showed strong soil preferences, for example Tripogon loliiformis for hard-setting soils and Dichanthium sericeum for clays. Common species, such as Chrysopogon fallax and Heteropogon contortus, had no strong soil preference. Wiregrasses (Aristida spp.) tended to be uncommon at both ends of the estimated pasture condition scale whereas H. contortus was far more common in pastures in good condition. Sedges (Cyperaceae) were common on all soil types and for all pasture condition ratings. Plants identified as increaser species were Tragus australianus, daisies (Asteraceae) and potentially toxic herbaceous legumes such as Indigofera spp. and Crotalaria spp. Pasture condition could not be reliably predicted based on the abundance of a single species or taxon but there may be scope for using integrated data for four to five ecologically contrasting plants such as Themeda triandra with daisies, T. loliiformis and flannel weeds (Malvaceae).

Optimising initial population density, growth time and nitrogen nutrition for assessing resistance of wheat cultivars to root-lesion nematode (Pratylenchus thornei)

Pratylenchus thornei is a major pathogen of wheat in Australia. Two glasshouse experiments with four wheat cultivars that had different final populations (Pf) of P. thornei in the field were used to optimise conditions for assessing resistance. With different initial populations (Pi) ranging up to 5250 P. thornei/kg soil, Pf of P. thornei increased to 16 weeks after sowing, and then decreased at 20 weeks in some cultivar x Pi combinations. The population dynamics of P. thornei up to 16 weeks were best described by a modified exponential equation P f (t) = aP i e kt where P f (t) is the final population density at time t, P i is the initial population density, a is the proportion of P i that initiates population development, and k is the intrinsic rate of increase of the population. The cultivar GS50a had very low k values at Pi of 5250 and 1050 indicating its resistance, Suneca and Potam had high k values indicating susceptibility, whereas intolerant Gatcher had a low value at the higher Pi and a high value at the lower Pi. Nitrate fertiliser increased plant growth and Pf values of susceptible cultivars, but in unplanted soil it decreased Pf. Nematicide (aldicarb 5 mg/kg soil) killed P. thornei more effectively in planted than in unplanted soil and increased plant growth particularly in the presence of N fertiliser. In both experiments, the wheat cultivars Suneca and Potam were more susceptible than the cultivar GS50a reflecting field results. The method chosen to discriminate wheat cultivars was to assess Pf after growth for 16 weeks in soil with Pi ~1050–5250 P. thornei/kg soil and fertilised with 200 mg NO3–N/kg soil.