Source fruits were then placed on top of the target fruits so that the inoculated stem end region would be in contact with the equatorial region of the target. Using the whole fruits was the most effective way to contact-inoculate oranges. Source fruits with visible mycelium and with visibly macerated tissue radially spread from the initial wound but no mycelium were placed on top of the equatorial side of three target oranges, with each infection site of the source orange in direct contact with one of the target oranges. The contact point between the source and target oranges was pre-wetted by spraying sterile water containing 0.01% Tween-20. Successful infections occurred when using oranges that did not have visible mycelium, while those with external mycelium failed to infect the target fruit. For tomatoes and apples, the contact inoculation procedure failed when using whole fruits due to leakage and accumulation of juices from the source fruits during incubation, leading to off-target infections in noncontact points. We then decided to use tissue sections instead of whole fruits to ensure a successful and uniform inoculation process. For tomatoes, a 3.5×3.5 cm, black plastic pots for plants square shaped pericarp section containing healthy/asymptomatic and symptomatic tissue in a 1:1 ratio was used.
Similarly, for apples, a cross-cut 1.5 cm thick was performed below and above the limit between symptomatic and asymptomatic tissue. The resulting disc was then cut into four 3×3×5.5 cm triangles containing decayed and macerated tissue in the middle and asymptomatic tissue on the sides. Source tissue sections of tomato and apple were placed on plastic boats, with the endocarp facing upwards for the tomato. Individual target tomatoes and apples with the equator side as the contact point were placed on top of their respective source tissue sections.The contact inoculation time was determined as the minimum time needed for successful disease development in the target fruits once the source fruits were detached. For control oranges and fungicide-treated oranges , the contact time was reduced to two days when using P. italicum WT and 1.5 days when using P. digitatum WT due to their advanced infection rates and aggressiveness on control and fungicide treated oranges. In all cases, contact inoculation was performed at room temperature, and target and source fruits were stored in high-humidity chambers .
Following the contact inoculation, the source whole fruits or tissue sections were removed, and target fruits were stored at 20 °C under high relative humidity until mycelium reached the equatorial region, or until evaluations were completed. Target oranges inoculated with P. italicum were stored at 20 °C for four days and at 15 °C for 12 days; while oranges inoculated with P. digitatum were stored at 15 °C for 15 days. Negative control samples were included for each trial. Control source fruits underwent the same steps as inoculated samples but were not inoculated with a fungal spore suspension.After contact inoculation, disease incidence, and severity were measured daily for tomatoes and every two days for apples and oranges. Disease incidence was calculated as the percentage of fruits displaying visual signs of tissue maceration or soft rot. Disease severity was obtained by measuring lesion area from pictures taken at each time point using a Nikon D5100 DSLR Camera with 18–55 mm f/3.5–5.6 and a custom-made macro in the ImageJ software.High-resolution multispectral images were taken using a VideometerLab 4 and processed with the VideometerLab software version 3.22.29. This equipment includes a sphere that uses strobe light-emitting diode technology to provide uniform and diffuse illumination. Reflectance images were taken at 19 wavelengths , including the long pass filters for a total of 50 spectral bands of the electromagnetic spectrum.
Multispectral images for all target fruits were taken with the stem end pointed to the side and the equator in the center before and after contact-based inoculation. For image analysis, pixels representing healthy and infected tissues were collected from a subset of fruit images from the total oranges. A normalized canonical discriminant analysis transformation based on the reflectance of each pixel was created to minimize the distance within classes and to maximize the distance among classes. A region of interest was obtained from all images by applying a mask to segment the fruits from the background. All fruits were collected in a blob database, and the healthy and infected areas were extracted based on the previously created nCDA transformation. Shape and spectral features were extracted from individual blobs, including area and tristimulus components of color, such as hue and saturation. The SpectralMean feature extracts the reflectance mean of each fruit for the 50 spectral bands. Region MSI_Mean calculates a trimmed mean of transformed pixel values within the blob , and MSIThreshold measures the percentage or area of the blob region with a transformation value higher than the threshold, based on the nCDA model .After choosing the appropriate source inoculum type and incubation periods, we tested the reliability and reproducibility of the contact-based inoculation method using different fungal species and strains on target fruits with and without fungicide treatments . In all the tests, the fungi spread from the contact points between the source and target fruit, causing tissue maceration and, in some cases, extensive mycelium growth . No symptoms were evident when non-infected fruits or tissues were used as source inoculum to control for secondary or unintended infections. Although we noticed some variation across different fruit commodities, fruit treatments, and fungal species and strains, most of our trials yielded disease incidence values of at least 80%, indicating the high performance of the contact-based inoculation method . Additionally, we observed that established lesions in the target fruits continued to expand over time, confirming that the fungal pathogens tested in this study could colonize and complete their life cycle once they penetrated the fruits . Wild-type strains of P. italicum and P. digitatum reached a disease incidence of 80% and 96.7% after 10 and 8 dpci, respectively, in untreated target oranges. In contrast, when using fungicide-treated oranges, a low overall disease incidence was observed for P. italicum and P. digitatum WT strains after 8 dpci and 14 dpci, respectively . On the other hand, disease incidence in fungicide and wax treated oranges contactinoculated with a P. digitatum fungicide-resistant strain was found to be 95.5%, which was similar to the untreated control oranges infected with P. digitatum WT . Lesion expansion in oranges varied depending on the target orange treatment and fungal isolate used . Overall, although lesion sizes were smaller for the FT oranges when using WT strains, the increase in lesion size across time points shows that the fungal pathogens tested in this study can colonize and complete their life cycle once they penetrate the target fruit. Field-grown tomatoes of the ‘Celebrity’ hybrid cultivar showed a high disease incidence, drainage pot reaching 90% across all evaluated time points when contact-inoculated with a B. cinerea WT strain . Meanwhile, other field grown tomatoes from the ‘Rutgers’ variety and ‘Shady Lady’ hybrid cultivar achieved the maximum disease incidence at 4 dpci and 6 dpci, with 80% and 94.4%, respectively. Commercial, greenhouse-grown, hybrid ‘Beefsteak’ tomatoes showed 100% disease incidence at 2 dpci when contact-inoculated with a B. cinerea FR strain. Tomatoes that did not show any disease incidence remained uninfected throughout the trial. Similarly to oranges, lesion size development varied across tomato cultivars and strains used, but shows the fungal strains used can colonize the target fruits . In apples, P. expansum FR infected 72.5% of the apples at 4 dpci and 90% at 6 dpci. The number of infected target apples remained unchanged until the last recorded time. Seven out of 40 apples never showed infection or lesion development from P. expansum and were considered as not infected during this study .
For the infected apples, lesion sizes steadily increased throughout the duration of the evaluation period .In oranges, Penicillium spp. growth showed minimal visual progression and seemed limited to the contact point until 12 dpci . However, using multispectral imaging we detected disease progression on the surface of target fruits, which was not apparent to the naked eye. Changes in the reflectance profiles of contact-inoculated oranges show that lesions appeared as early as 8 dpci and continued to expand even when mature lesions and visible mycelium were only obvious until 14 dpci and 18 dpci, respectively . On the other hand, practically no changes in the reflectance profile were observed for control fruits . Furthermore, normalized canonical discriminant analysis transformation, which combines all wavelengths, could detect changes in areas where lesions were to be developed in earlier time points. These areas were calculated using a threshold in the nCDA transformed scale ranging from healthy tissues to infected tissues . The highest separation potential in the nCDA transformation was obtained when combining about ten wavelengths with minor gain with additional wavelengths . Using MSI to monitor pathogen growth revealed the early onset of infection in oranges. It also can be used for sensitive quantification of lesion area before the disease is visible to the naked eye.Consistent and reliable inoculation methods that mimic natural conditions and industry scenarios are key for the study of plant-pathogen interactions and the development of post harvest control methods. Here, we established a non-wounding, contact inoculation protocol that recreates the infections naturally occurring in the post harvest supply chain through nesting. We produced whole fruits or tissues that were infected and served as inoculum sources for spreading the disease to healthy fruit. In all fruits, except for fungicide-treated oranges contact inoculated with wild-type pathogen strains, disease incidence rates of 80% or higher were observed by the final evaluation time point, showing the effectiveness of the proposed methodology. Disease severity measurements were used to evaluate disease progression and growth behavior of fungal pathogens, confirming successful infections beyond the initial contact point in each fruit commodity. Even though disease incidence was lower in fungicide-treated oranges contact-inoculated with fungicide-sensitive pathogens, having successful infections in fungicide-treated oranges showed that the pathogens were capable of causing disease through this protocol, even in disadvantageous conditions. Furthermore, visualizing the fruits using MSI allowed us to confirm that, although lesions were not visible to the naked eye until 14 days post-contact inoculation with P. digitatum, the fungus was established in the fruit tissue and growing by 8 dpci. While successful B. cinerea infections were observed in all tomato trials, differences in disease incidence and severity values were observed across varieties. ‘Rutgers’ tomatoes exhibited greater variation between infection rates and a lower disease incidence than other tomato varieties. These differences could be due to surface characteristics, such as cuticle thickness and permeability, which influence the generation of pathogen-induced signals that activate defense responses. The ‘Rutgers’ tomatoes are more similar to processing types. Furthermore, thicker fruit cuticles provide higher resistance to initial B. cinerea infections. Also, differences in pathogen behavior were observed in this study, particularly between Penicillium spp. when contact inoculated in oranges and apples. P. italicum WT showed an earlier mycelium appearance as compared to P. digitatum WT, although it was the latter that covered the fruits completely in mycelium first. In apples, P. expansum showed a slower disease incidence and severity progression than the other commodities. This could be due to differences in infection mechanisms between fungal species, as well as due to the fact that blue mold develops better at cooler temperatures compared to green mold. For all three fruit commodities tested, a percentage of the fruits did not get infected, even after several weeks of monitoring. This could be because the fruit was able to halt pathogen infection at the contact point, preventing it from spreading throughout the tissues. Another possibility is that, despite homogeneous incubation conditions, fruit-specific microclimates were not always conducive to disease development in all fruits, even though we ensured consistency of the technical aspects of the protocol. The overall results of this study confirm that the non-wounding, contact-based inoculation method was effective in all fruit-pathogen interactions tested. Although an initial step of wound-inoculating the source fruit material is required, spread of the disease to healthy target fruits is done solely through contact between the tissues, and the target fruits remain unwounded throughout the entire procedure. This method holds promise for further application in other pathosystems by focusing on several key aspects. First, it is crucial to recognize the importance of the homogeneity of the fruits used as source inoculum and the initial 24 h during contact inoculation, as both play a significant role in the establishment and spread of fungal infections. For example, it is recommended that, if possible, fruits should come from the same location and supplier, and transportation-storage conditions should remain constant. Also, the incubation should be done at high humidity with some level of gas exchange , especially after the first day of contact. Second, investigating the contact time between source and target fruits is essential to ensure the accuracy and reproducibility of the inoculation method. Third, exploring the position of infected fruits or tissues, which serves as the source of inoculation, will help identify the most favorable conditions for efficient pathogen transfer between fruits through contact.