Limits...
Plant ionome diagnosis using sound balances: case study with mango (Mangifera Indica).

Parent SÉ, Parent LE, Rozane DE, Natale W - Front Plant Sci (2013)

Bottom Line: Traditional multivariate methods were found to be biased.A customized receiver operating characteristic (ROC) iterative procedure was developed to classify tissue ionomes between balanced/misbalanced and high/low-yielders.The ROC partitioning procedure showed that the critical Mahalanobis distance of 4.08 separating balanced from imbalanced specimens about yield cut-off of 128.5 kg fruit tree(-1) proved to be a fairly informative test (area under curve = 0.84-0.92).

View Article: PubMed Central - PubMed

Affiliation: Department of Soils and Agrifood Engineering, ERSAM, Université Laval Québec, QC, Canada.

ABSTRACT
Plant ionomes and soil nutrients are commonly diagnosed in agronomy using concentration and nutrient ratio ranges. However, both diagnoses are biased by redundancy of information, subcompositional incoherence and non-normal distribution inherent to compositional data, potentially leading to conflicting results and wrong inferences. Our objective was to present an unbiased statistical approach of plant nutrient diagnosis using a balance concept and mango (Mangifera indica) as test crop. We collected foliar samples at flowering stage in 175 mango orchards. The ionomes comprised 11 nutrients (S, N, P, K, Ca, Mg, B, Cu, Zn, Mn, Fe). Traditional multivariate methods were found to be biased. Ionomes were thus represented by unbiased balances computed as isometric log ratios (ilr). Soil fertility attributes (pH and bioavailable nutrients) were transformed into balances to conduct discriminant analysis. The orchards differed more from genotype than soil nutrient signatures. A customized receiver operating characteristic (ROC) iterative procedure was developed to classify tissue ionomes between balanced/misbalanced and high/low-yielders. The ROC partitioning procedure showed that the critical Mahalanobis distance of 4.08 separating balanced from imbalanced specimens about yield cut-off of 128.5 kg fruit tree(-1) proved to be a fairly informative test (area under curve = 0.84-0.92). The [P / N,S] and [Mn / Cu,Zn] balances were found to be potential sources of misbalance in the less productive orchards, and should thus be further investigated in field experiments. We propose using a coherent pan balance diagnostic method with median ilr values of top yielders centered at fulcrums of a mobile and the critical Mahalanobis distance as a guide for global nutrient balance. Nutrient concentrations in weighing pans assisted appreciating nutrients as relative shortage, adequacy or excess in balances.

No MeSH data available.


Related in: MedlinePlus

Binary classification of data with indexes (top). Class sum of squares (bottom).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3824108&req=5

Figure 3: Binary classification of data with indexes (top). Class sum of squares (bottom).

Mentions: The area under the ROC curve (AUC), computed by summing rectangles under the step curve, reached a peak at 0.84 (Figure 2B), a value comparable to the AUC for fairly informative tests (0.80–0.98) in medical sciences (Swets, 1988). The ROC curve did not show a monotonic decrease of sensitivity as specificity increased, as usually observed in ROC diagnoses due to the re-sampling of the TN specimens (see methodology section), which is generally not needed in conventional clinical studies. The AUC computed across the fitted binormal model (Hanley, 1988) returned a value of 0.89, ranging between 0.84 and 0.92, with a confidence level of 95% (Figure 2B). The response delimiter corresponding to the AUC peak was 128.5 kg fresh fruit tree−1 (Figure 2A). The optimal compromise between specificity and sensitivity for the optimal response delimiter was found at a specificity of 0.95 and a sensitivity of 0.92, corresponding to a predictor threshold (Mahalanobis distance) of 4.08 (Figure 2B). Results of the binary classification are presented in Figure 3, where the two optimal delimiters classified 20 observations in the TN quadrants. The semi-transparent ellipse enclosed 95% of the theoretical distribution of all observations. The TN group was essentially constituted of “Tommy” (6) and “Palmer” (14) orchards. All “Espada” and “Haden” orchards were classified as TP. The large majority (accuracy = 92%) of specimens were correctly diagnosed by the Mahalanobis distance predictor. Almost all specimens declared imbalanced yielded less than cut-off yield value (PPV = 99%). On the other hand, nearly two thirds (NPV = 65%) of balanced specimens yielded more than cut-off yield value. Median ilr values of TN specimens as well as the covariance matrix used to measure Mahalanobis distances are presented in Tables 3 and 4.


Plant ionome diagnosis using sound balances: case study with mango (Mangifera Indica).

Parent SÉ, Parent LE, Rozane DE, Natale W - Front Plant Sci (2013)

Binary classification of data with indexes (top). Class sum of squares (bottom).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3824108&req=5

Figure 3: Binary classification of data with indexes (top). Class sum of squares (bottom).
Mentions: The area under the ROC curve (AUC), computed by summing rectangles under the step curve, reached a peak at 0.84 (Figure 2B), a value comparable to the AUC for fairly informative tests (0.80–0.98) in medical sciences (Swets, 1988). The ROC curve did not show a monotonic decrease of sensitivity as specificity increased, as usually observed in ROC diagnoses due to the re-sampling of the TN specimens (see methodology section), which is generally not needed in conventional clinical studies. The AUC computed across the fitted binormal model (Hanley, 1988) returned a value of 0.89, ranging between 0.84 and 0.92, with a confidence level of 95% (Figure 2B). The response delimiter corresponding to the AUC peak was 128.5 kg fresh fruit tree−1 (Figure 2A). The optimal compromise between specificity and sensitivity for the optimal response delimiter was found at a specificity of 0.95 and a sensitivity of 0.92, corresponding to a predictor threshold (Mahalanobis distance) of 4.08 (Figure 2B). Results of the binary classification are presented in Figure 3, where the two optimal delimiters classified 20 observations in the TN quadrants. The semi-transparent ellipse enclosed 95% of the theoretical distribution of all observations. The TN group was essentially constituted of “Tommy” (6) and “Palmer” (14) orchards. All “Espada” and “Haden” orchards were classified as TP. The large majority (accuracy = 92%) of specimens were correctly diagnosed by the Mahalanobis distance predictor. Almost all specimens declared imbalanced yielded less than cut-off yield value (PPV = 99%). On the other hand, nearly two thirds (NPV = 65%) of balanced specimens yielded more than cut-off yield value. Median ilr values of TN specimens as well as the covariance matrix used to measure Mahalanobis distances are presented in Tables 3 and 4.

Bottom Line: Traditional multivariate methods were found to be biased.A customized receiver operating characteristic (ROC) iterative procedure was developed to classify tissue ionomes between balanced/misbalanced and high/low-yielders.The ROC partitioning procedure showed that the critical Mahalanobis distance of 4.08 separating balanced from imbalanced specimens about yield cut-off of 128.5 kg fruit tree(-1) proved to be a fairly informative test (area under curve = 0.84-0.92).

View Article: PubMed Central - PubMed

Affiliation: Department of Soils and Agrifood Engineering, ERSAM, Université Laval Québec, QC, Canada.

ABSTRACT
Plant ionomes and soil nutrients are commonly diagnosed in agronomy using concentration and nutrient ratio ranges. However, both diagnoses are biased by redundancy of information, subcompositional incoherence and non-normal distribution inherent to compositional data, potentially leading to conflicting results and wrong inferences. Our objective was to present an unbiased statistical approach of plant nutrient diagnosis using a balance concept and mango (Mangifera indica) as test crop. We collected foliar samples at flowering stage in 175 mango orchards. The ionomes comprised 11 nutrients (S, N, P, K, Ca, Mg, B, Cu, Zn, Mn, Fe). Traditional multivariate methods were found to be biased. Ionomes were thus represented by unbiased balances computed as isometric log ratios (ilr). Soil fertility attributes (pH and bioavailable nutrients) were transformed into balances to conduct discriminant analysis. The orchards differed more from genotype than soil nutrient signatures. A customized receiver operating characteristic (ROC) iterative procedure was developed to classify tissue ionomes between balanced/misbalanced and high/low-yielders. The ROC partitioning procedure showed that the critical Mahalanobis distance of 4.08 separating balanced from imbalanced specimens about yield cut-off of 128.5 kg fruit tree(-1) proved to be a fairly informative test (area under curve = 0.84-0.92). The [P / N,S] and [Mn / Cu,Zn] balances were found to be potential sources of misbalance in the less productive orchards, and should thus be further investigated in field experiments. We propose using a coherent pan balance diagnostic method with median ilr values of top yielders centered at fulcrums of a mobile and the critical Mahalanobis distance as a guide for global nutrient balance. Nutrient concentrations in weighing pans assisted appreciating nutrients as relative shortage, adequacy or excess in balances.

No MeSH data available.


Related in: MedlinePlus