Robustness of NIR calibrations for assessing fruit quality

Predictive models based on near infra-red spectroscopy for the assessment of fruit internal quality attributes must exhibit a degree of robustness across the parameters of variety, district and time to be of practical use in fruit grading. At the time this thesis was initiated, while there were a number of published reports on the development of near infra-red based calibration models for the assessment of internal quality attributes of intact fruit, there were no reports of the reliability ("robustness") of such models across time, cultivars or growing regions. As existing published reports varied in instrumentation employed, a re-analysis of existing data was not possible. An instrument platform, based on partial transmittance optics, a halogen light source and (Zeiss MMS 1) detector operating in the short wavelength near infra-red region was developed for use in the assessment of intact fruit. This platform was used to assess populations of macadamia kernels, melons and mandarin fruit for total soluble solids, dry matter and oil concentration. Calibration procedures were optimised and robustness assessed across growing areas, time of harvest, season and variety. In general, global modified partial least squares regression (MPLS) calibration models based on derivatised absorbance data were better than either multiple linear regression or `local' MPLS models in the prediction of independent validation populations . Robustness was most affected by growing season, relative to the growing district or variety . Various calibration updating procedures were evaluated in terms of calibration robustness. Random selection of samples from the validation population for addition to the calibration population was equivalent to or better than other methods of sample addition (methods based on the Mahalanobis distance of samples from either the centroid of the population or neighbourhood samples). In these exercises the global Mahalanobis distance (GH) was calculated using the scores and loadings from the calibration population on the independent validation population. In practice, it is recommended that model predictive performance be monitored in terms of predicted sample GH, with model updating using as few as 10 samples from the new population undertaken when the average GH value exceeds 1 .0 .

Effects of seasonal variability on FT-NIR prediction of dry matter content for whole Hass avocado fruit

Fourier Transform (FT)-near infra-red spectroscopy (NIRS) was investigated as a non-invasive technique for estimating percentage (%) dry matter of whole intact 'Hass' avocado fruit. Partial least squares (PLS) calibration models were developed from the diffuse reflectance spectra to predict % dry matter, taking into account effects of seasonal variation. It is found that seasonal variability has a significant effect on model predictive performance for dry matter in avocados. The robustness of the calibration model, which in general limits the application for the technique, was found to increase across years (seasons) when more seasonal variability was included in the calibration set. The R-v(2) and RMSEP for the single season prediction models predicting on an independent season ranged from 0.09 to 0.61 and 2.63 to 5.00, respectively, while for the two season models predicting on the third independent season, they ranged from 0.34 to 0.79 and 2.18 to 2.50, respectively. The bias for single season models predicting an independent season was as high as 4.429 but <= 1.417 for the two season combined models. The calibration model encompassing fruit from three consecutive years yielded predictive statistics of R-v(2) = 0.89, RMSEP = 1.43% dry matter with a bias of -0.021 in the range 16.1-39.7% dry matter for the validation population encompassing independent fruit from the three consecutive years. Relevant spectral information for all calibration models was obtained primarily from oil, carbohydrate and water absorbance bands clustered in the 890-980, 1005-1050, 1330-1380 and 1700-1790 nm regions. These results indicate the potential of FT-NIRS, in diffuse reflectance mode, to non-invasively predict the % dry matter of whole 'Hass' avocado fruit and the importance of the development of a calibration model that incorporates seasonal variation. Crown Copyright (c) 2012 Published by Elsevier B.V. All rights reserved.

Projecting future expansion of invasive species: Comparing and improving methodologies for species distribution modeling

Modeling the distributions of species, especially of invasive species in non-native ranges, involves multiple challenges. Here, we developed some novel approaches to species distribution modeling aimed at reducing the influences of such challenges and improving the realism of projections. We estimated species-environment relationships with four modeling methods run with multiple scenarios of (1) sources of occurrences and geographically isolated background ranges for absences, (2) approaches to drawing background (absence) points, and (3) alternate sets of predictor variables. We further tested various quantitative metrics of model evaluation against biological insight. Model projections were very sensitive to the choice of training dataset. Model accuracy was much improved by using a global dataset for model training, rather than restricting data input to the species’ native range. AUC score was a poor metric for model evaluation and, if used alone, was not a useful criterion for assessing model performance. Projections away from the sampled space (i.e. into areas of potential future invasion) were very different depending on the modeling methods used, raising questions about the reliability of ensemble projections. Generalized linear models gave very unrealistic projections far away from the training region. Models that efficiently fit the dominant pattern, but exclude highly local patterns in the dataset and capture interactions as they appear in data (e.g. boosted regression trees), improved generalization of the models. Biological knowledge of the species and its distribution was important in refining choices about the best set of projections. A post-hoc test conducted on a new Partenium dataset from Nepal validated excellent predictive performance of our “best” model. We showed that vast stretches of currently uninvaded geographic areas on multiple continents harbor highly suitable habitats for Parthenium hysterophorus L. (Asteraceae; parthenium). However, discrepancies between model predictions and parthenium invasion in Australia indicate successful management for this globally significant weed. This article is protected by copyright. All rights reserved.

Projecting future expansion of invasive species: Comparing and improving methodologies for species distribution modeling

Modeling the distributions of species, especially of invasive species in non-native ranges, involves multiple challenges. Here, we developed some novel approaches to species distribution modeling aimed at reducing the influences of such challenges and improving the realism of projections. We estimated species-environment relationships with four modeling methods run with multiple scenarios of (1) sources of occurrences and geographically isolated background ranges for absences, (2) approaches to drawing background (absence) points, and (3) alternate sets of predictor variables. We further tested various quantitative metrics of model evaluation against biological insight. Model projections were very sensitive to the choice of training dataset. Model accuracy was much improved by using a global dataset for model training, rather than restricting data input to the species’ native range. AUC score was a poor metric for model evaluation and, if used alone, was not a useful criterion for assessing model performance. Projections away from the sampled space (i.e. into areas of potential future invasion) were very different depending on the modeling methods used, raising questions about the reliability of ensemble projections. Generalized linear models gave very unrealistic projections far away from the training region. Models that efficiently fit the dominant pattern, but exclude highly local patterns in the dataset and capture interactions as they appear in data (e.g. boosted regression trees), improved generalization of the models. Biological knowledge of the species and its distribution was important in refining choices about the best set of projections. A post-hoc test conducted on a new Partenium dataset from Nepal validated excellent predictive performance of our “best” model. We showed that vast stretches of currently uninvaded geographic areas on multiple continents harbor highly suitable habitats for Parthenium hysterophorus L. (Asteraceae; parthenium). However, discrepancies between model predictions and parthenium invasion in Australia indicate successful management for this globally significant weed. This article is protected by copyright. All rights reserved.

Evaluating the performance of weed containment programmes

Aim: To develop approaches to the evaluation of programmes whose strategic objectives are to halt or slow weed spread. Location: Australia. Methods: Key aspects in the evaluation of weed containment programmes are considered. These include the relevance of models that predict the effects of management intervention on spread, the detection of spread, evidence for containment failure and metrics for absolute or partial containment. Case studies documenting either near-absolute (Orobanche ramosa L., branched broomrape) or partial (Parthenium hysterophorus (L.) King and Robinson, parthenium) containment are presented. Results: While useful for informing containment strategies, predictive models cannot be employed in containment programme evaluation owing to the highly stochastic nature of realized weed spread. The quality of observations is critical to the timely detection of weed spread. Effectiveness of surveillance and monitoring activities will be improved by utilizing information on habitat suitability and identification of sites from which spread could most compromise containment. Proof of containment failure may be difficult to obtain. The default option of assuming that a new detection represents containment failure could lead to an underestimate of containment success, the magnitude of which will depend on how often this assumption is made. Main conclusions: Evaluation of weed containment programmes will be relatively straightforward if containment is either absolute or near-absolute and may be based on total containment area and direct measures of containment failure, for example, levels of dispersal, establishment and reproduction beyond (but proximal to) the containment line. Where containment is only partial, other measures of containment effectiveness will be required. These may include changes in the rates of detection of new infestations following the institution of interventions designed to reduce dispersal, the degree of compliance with such interventions, and the effectiveness of tactics intended to reduce fecundity or other demographic drivers of spread. © 2012 Blackwell Publishing Ltd.

Interaction between production characteristics and postharvest performance and practices for fresh fruit

Significant interactions have been demonstrated between production factors and postharvest quality of fresh fruit. Accordingly, there is an attendant need for adaptive postharvest actions to modulate preharvest effects. The most significant preharvest effects appear to be mediated through mineral nutrition influences on the physical characteristics of fruit. Examples of specific influencers include fertilisers, water availability, rootstock, and crop load effects on fruit quality attributes such as skin colour, susceptibility to diseases and physiological disorders, and fruit nutritional composition. Also, rainfall before and during harvest can markedly affect fruit susceptibility to skin blemishes, physical damage, and diseases. Knowledge of preharvest-postharvest interactions can help determine the basis for variability in postharvest performance and thereby allow refinement of postharvest practices to minimise quality loss after harvest. This knowledge can be utilised in predictive management systems. Such systems can benefit from characterisation of fruit nutritional status, particularly minerals, several months before and/or at harvest to allow informed decisions on postharvest handling and marketing options. Other examples of proactive management practices include adjusting harvesting and packing systems to account for rainfall effects before and/or during harvest. Improved understanding of preharvest-postharvest interactions is contributing to the delivery of consistently higher quality of fruit to consumers. This paper focuses on the state of knowledge for sub-tropical and tropical fruits, in particular avocado and mango.

Interaction between production characteristics and postharvest performance and practices for fresh fruit

Significant interactions have been demonstrated between production factors and postharvest quality of fresh fruit. Accordingly, there is an attendant need for adaptive postharvest actions to modulate preharvest effects. The most significant preharvest effects appear to be mediated through mineral nutrition influences on the physical characteristics of fruit. Examples of specific influencers include fertilisers, water availability, rootstock, and crop load effects on fruit quality attributes such as skin colour, susceptibility to diseases and physiological disorders, and fruit nutritional composition. Also, rainfall before and during harvest can markedly affect fruit susceptibility to skin blemishes, physical damage, and diseases. Knowledge of preharvest-postharvest interactions can help determine the basis for variability in postharvest performance and thereby allow refinement of postharvest practices to minimise quality loss after harvest. This knowledge can be utilised in predictive management systems. Such systems can benefit from characterisation of fruit nutritional status, particularly minerals, several months before and/or at harvest to allow informed decisions on postharvest handling and marketing options. Other examples of proactive management practices include adjusting harvesting and packing systems to account for rainfall effects before and/or during harvest. Improved understanding of preharvest-postharvest interactions is contributing to the delivery of consistently higher quality of fruit to consumers. This paper focuses on the state of knowledge for sub-tropical and tropical fruits, in particular avocado and mango.