https://doi.org/10.3389/fpls.2022.1033308
“Bitter pit (BP) is one of the most relevant post-harvest disorders for apple industry worldwide, which is often related to calcium (Ca) deficiency at the calyx end of the fruit. Its occurrence takes place along with an imbalance with other minerals, such as potassium (K). Although the K/Ca ratio is considered a valuable indicator of BP, a high variability in the levels of these elements occurs within the fruit, between fruits of the same plant, and between plants and orchards. Prediction systems based on the content of elements in fruit have a high variability because they are determined in samples composed of various fruits. With X-ray fluorescence (XRF) spectrometry, it is possible to characterize non-destructively the signal intensity for several mineral elements at a given position in individual fruit and thus, the complete signal of the mineral composition can be used to perform a predictive model to determine the incidence of bitter pit. Therefore, it was hypothesized that using a multivariate modeling approach, other elements beyond the K and Ca could be found that could improve the current clutter prediction capability. Two studies were carried out: on the first one an experiment was conducted to determine the K/Ca and the whole spectrum using XRF of a balanced sample of affected and non-affected ‘Granny Smith’ apples. On the second study apples of three cultivars (‘Granny Smith’, ‘Brookfield’ and ‘Fuji’), were harvested from two commercial orchards to evaluate the use of XRF to predict BP. With data from the first study a multivariate classification system was trained (balanced database of healthy and BP fruit, consisting in 176 from each group) and then the model was applied on the second study to fruit from two orchards with a history of BP. Results show that when dimensionality reduction was performed on the XRF spectra (1.5 – 8 KeV) of ‘Granny Smith’ apples, comparing fruit with and without BP, along with K and Ca, four other elements (i.e., Cl, Si, P, and S) were found to be deterministic. However, the PCA revealed that the classification between samples (BP vs. non-BP fruit) was not possible by univariate analysis (individual elements or the K/Ca ratio).Therefore, a multivariate classification approach was applied, and the classification measures (sensitivity, specificity, and balanced precision) of the PLS-DA models for all cultivars evaluated (‘Granny Smith’, ‘Fuji’ and ‘Brookfield’) on the full training samples and with both validation procedures (Venetian and Monte Carlo), ranged from 0.76 to 0.92. The results of this work indicate that using this technology at the individual fruit level is essential to understand the factors that determine this disorder and can improve BP prediction of intact fruit.”
”
2.1 Apple fruit used for bitter pit prediction
Two studies were carried out as follows:
2.1.1 Study 1
A balanced database was generated with apples with and without BP symptoms. For this purpose, during the 2018/19 season, Dole Chile Co. provided two groups of ‘Granny Smith’ apples from a commercial lot of refrigerated storage (5 months at 0°C and 90% RH): i) lot 1 with BP: 176 apples with medium and severe epidermal incidence of BP (i.e., 3 – 5, and >5 pits, respectively); and ii) lot 2 without BP: 176 apples with no evidence of BP injury (i.e., no epidermal indication of BP or other disorder). Since most of the BP damage is concentrated towards the distal region of the fruit, the evaluations were concentrated in the calyx area (Figure 1A-1). To determine the variability associated with XRF measurements at fruit level, a reproducibility study was conducted, in which six equidistant points were measured along the calyx end of each pitted and non-pitted fruit (Figure 1A-2).
FIGURE 1 Section of the fruit calyx (dashed red line) where the bitter pit is usually located (A-1) and where the reproducibility XRF study was performed at six equidistant points (A-2); details of XRF measurement (B); bitter pit in fruit with no external symptoms, before (C-1) and after fruit peeling (C-2).
”
”
3.1 Study 1: Determination of the K/Ca ratio by XRF on fruit with and without bitter pit
When the balanced data base of ‘Granny Smith’ apples with and without BP was studied, the X-ray spectra (Figure 2) showed that the most significant peaks corresponded to potassium (K: 3.31 and 3.59 keV), calcium (Ca: 3.69 and 4.01 keV), and chloride (Cl: 2.68 and 2.82 keV); less intense peaks were found for silicon (Si: 2.02 and 2.46 keV), phosphorus (P: 2.02 and 2.14 keV), and sulfur (S: 2.31 and 2.46 keV). For these spectra, the most affected samples exhibited higher K intensity, while non-affected apples showed higher Ca intensity (Figure 2).
FIGURE 2 General XRF spectra for ‘Granny Smith’ apple epidermis for fruit with (red) and without (blue) bitter pit; K and Ca peaks details in small graphic. Similar spectra were found in the other cultivars.
Effectively, the analyzed spectra by a PCA model showed that K (3.31 keV) and Ca (3.69 keV) peaks contributed with the maxima variability for each component (Figure 3A). The dispersion of these two signals, evaluated as standard deviation and coefficient of variation, are shown in Supplementary Table 1. The results evidenced that the random error for K and Ca responses varied in the range of 10 to 25% (data not shown); considering that instrument variation is lower than 10%, this dispersion could be associated to heterogeneity within the apple composition. Additionally, the score plot (Figure 3B) obtained after PCA analysis shows a severe overlapping between samples with and without BP, suggesting an important similarity between spectral signature of the fruits.
FIGURE 3 Loading (A) and score (B) plots obtained with XRF-spectra of ‘Granny Smith’ apples with (blue) and without (red) bitter pit.
Especially for the K/Ca ratio calculated after the element deconvolution of the data matrix of ‘Granny Smith’, the results showed a high variability in sound fruits (~5 – 23), differences that almost doubled in the case of affected ones (the ratio ranged from ~4 – 41) (Supplementary Table 1). Globally, 80% of the fruit without BP had a K/Ca ratio of less than ~7, while in those with BP, for the same ratio, the proportion was less than 24% (Figure 4A); although with different scales, similar results were found for K/Ca (Supplementary Table 1, and Figure 4B). From same data base, the analysis of variance (Supplementary Table 1) the cumulative frequency distribution and the box-and-whisker plots (Figures 4C–H) of all potential identified elements indicated that, except for Cl and partially for P (p=0.0931), the rest of the elements differed between BP groups (p≤ 0.00001). On the other hand, from a methodological point of view, no differences were found among the six equidistant points at the calyx-end for any of the elements demonstrating that measurements at the calyx is consistent. Furthermore, no significant interactions were found between the two factors, indicating that the main difference between samples was due to the BP presence/absence (Supplementary Table 1). K/Ca ratio significance was the same regardless of the way in which they were calculated (deconvolution method vs direct ratio from maximum K and Ca signals); also, a high association (R2 = 0.98) was found between both (Supplementary Figure 1).
FIGURE 4 Comparison of the cumulative frequency distributions and box-and-whisker plots of K/Ca ratio calculated by the deconvolution method (decon) (A) and direct ratio (dr) (B) from maximum K and Ca signals, and the signal insensitivity of the mineral elements with the greatest preponderance in ‘Granny Smith’ fruit with (red) or without (green) bitter pit: Si (C), P (D), S (E), Cl (F), K (G), and Ca (H).
The best PCA models were obtained after mean-centering and using 4 or 5 components to reach explained variances higher than 90%. As observed in Figure 3, the score plot resulted in 69.8% of explained variance for PC1 and 14.1% for PC2 (Figure 3A) and ~90% when PC3 was included. The more explicative spectral region considers the characteristic region of K, Ca, and Cl (i.e., 2.6 to 3.7 keV). Nevertheless, a significant influence is observed for lower energy signals between 1.7 and 2.5 keV, typical for Si, S, and P.
“